adopting a label: heterogeneity in the economic consequences around ias/ifrs adoptions

DOI: 10.1111/1475-679X.12005Journal of Accounting Research

Vol. 51 No. 3 June 2013Printed in U.S.A.

Adopting a Label: Heterogeneity inthe Economic ConsequencesAround IAS/IFRS Adoptions

H O L G E R D A S K E , ∗ L U Z I H A I L , † C H R I S T I A N L E U Z , ‡A N D R O D R I G O V E R D I §

Received 5 October 2011; accepted 13 December 2012

ABSTRACT

This study examines liquidity and cost of capital effects around voluntaryand mandatory IAS/IFRS adoptions. In contrast to prior work, we focus onthe firm-level heterogeneity in the economic consequences, recognizing thatfirms have considerable discretion in how they implement the new standards.Some firms may make very few changes and adopt IAS/IFRS more in name,while for others the change in standards could be part of a strategy to increasetheir commitment to transparency. To test these predictions, we classify firmsinto “label” and “serious” adopters using firm-level changes in reporting in-centives, actual reporting behavior, and the external reporting environmentaround the switch to IAS/IFRS. We analyze whether capital-market effects are

∗University of Mannheim; †The Wharton School, University of Pennsylvania; ‡The Univer-sity of Chicago Booth School of Business and NBER; §Sloan School of Management, MIT.

We appreciate the helpful comments of an anonymous referee, Ray Ball, Hans Christensen,Joachim Gassen, Gunther Gebhardt, Bob Holthausen, Bjorn Jorgensen, Peter Wysocki, SteveYoung, and workshop participants at the 2007 Utah Winter Accounting Conference, 2007Harvard Business School IMO Conference, 2007 Global Issues in Accounting Conference,2007 Journal of Accounting, Auditing and Finance Conference, 2008 AAA Annual Meet-ing, University of Chicago, INSEAD, Lancaster University, University of Lugano, University ofMannheim, Massachusetts Institute of Technology, University of Missouri, and New York Uni-versity. We thank Eric Blouin, Emily Goergen, Wannia Hu, Yuji Maruyama, Anindya Mishra,Michael Sall, Caleb Smith, Richie Wan, and Kai Wright for their excellent research assistance.The data set of voluntary IAS reporting created for this study is available for download here:http://research.chicagobooth.edu/arc/journal/onlineappendices.aspx.

495

Copyright C©, University of Chicago on behalf of the Accounting Research Center, 2013

496 H. DASKE, L. HAIL, C. LEUZ, AND R. VERDI

different across “serious” and “label” firms. While on average liquidity andcost of capital often do not change around voluntary IAS/IFRS adoptions,we find considerable heterogeneity: “Serious” adoptions are associated withan increase in liquidity and a decline in cost of capital, whereas “label” adop-tions are not. We obtain similar results when classifying firms around manda-tory IFRS adoption. Our findings imply that we have to exercise caution wheninterpreting capital-market effects around IAS/IFRS adoption as they also re-flect changes in reporting incentives or in firms’ broader reporting strategies,and not just the standards.

1. Introduction

Over the last number of years the adoption of International Financial Re-porting Standards (IFRS) or, before that, International Accounting Stan-dards (IAS) has gained considerable momentum around the world.1 Over100 countries require or permit the use of IFRS for financial reporting pur-poses, and several more are considering a mandate or at least giving firmsan option to report under IFRS. Even before the IFRS mandate, many firmshave voluntarily adopted IAS. An important issue for the literature analyz-ing capital-market outcomes around IAS and IFRS adoptions is whether theobserved consequences can be attributed to the change in standards itself,and hence can be labeled IAS or IFRS effects.

To highlight the difficulties in isolating such effects, we examine cross-sectional differences in capital-market outcomes around IAS/IFRS adop-tions and the role of firm-level reporting incentives in explaining them. Weshow that a simple partitioning based on changes in firm-level reportingincentives around IAS/IFRS adoptions has significant explanatory powerfor the direction and magnitude of the observed capital-market outcomes.The evidence is consistent with pervasive selection effects, particularly forvoluntary adoptions, and cautions us to interpret observed capital-marketoutcomes as a response to the change in standards alone. The evidence alsoquestions the extent to which a switch to IAS/IFRS by itself (without anyother changes) represents a commitment to a certain level of transparency.

For the most part, our analyses center on voluntary IAS adoptions. We doso for three reasons. First, our research design relies crucially on having suf-ficient variation in the underlying motives for IAS adoption to illustrate theselection issue. In this regard, studying voluntary adopters helps to increasethe power of the tests. Yet, as we demonstrate later in the paper, reportingincentives also play a role around mandatory IFRS adoption, for instance,in determining whether firms resist changing their reporting practices. Sec-ond, voluntary IAS adoptions are not as clustered in time as mandatoryadoptions but rather spread over multiple periods (from 1990 to 2005 in

1 See, e.g., Barth [2008] and Hail, Leuz, and Wysocki [2010]. In 2001, IAS were renamedto IFRS. In this paper, we use the two terms interchangeably, but primarily refer to IAS as oursample period goes back as far as 1990.

ADOPTING A LABEL 497

our sample). This feature makes it easier to eliminate the effects of un-related economic shocks and institutional changes. Third, understandingthe effects around voluntary adoptions is useful even in times when IFRSreporting is mandatory in many countries. For instance, firms that adoptedIAS/IFRS voluntarily can serve as an important benchmark in studies thatseek to identify the effects of mandatory IFRS reporting (e.g., Christensen,Hail, and Leuz [2012]). Furthermore, voluntary adoption is relevant to pri-vate firms that in most jurisdictions are not subject to the IFRS mandateand to firms with the option to adopt IFRS in countries not yet imposing amandate.

Our main hypothesis is that the observed economic consequencesaround IAS adoptions depend on management’s reporting incentives, in-cluding the underlying motivations for the accounting change, rather thanthe change in accounting standards per se. As a result we expect predictableheterogeneity in outcomes across firms. In particular, one frequently voicedconcern is that some firms adopt IAS merely in name without making ma-terial changes to their reporting policies (e.g., Ball [2001, 2006]). We referto these firms as “label adopters.” Firms may also adopt IAS as part of abroader strategy to increase their commitment to transparency. We refer tothese firms as “serious adopters,” meaning that they are “serious” about thechange in their reporting strategy (which includes but is not necessarily lim-ited to IAS adoption). Our study is designed to identify and examine suchdifferences across firms. Provided that investors can differentiate between“serious” and “label” firms, we should observe differential capital-marketeffects, e.g., in market liquidity and cost of capital.

To conduct our analysis, we construct a large panel of voluntary IAS adop-tions from 1990 to 2005 across 30 countries. We identify voluntary IAS adop-tions based on accounting standards data in Worldscope, Global Vantage,and an extensive hand-collection of more than 20,000 annual reports. Wecreate an IAS reporting panel, which allows us to provide descriptive ev-idence on the adoption trends around the world as well as on firms’ in-dividual adoption strategies (see the appendix). As expected, the numberof firms reporting under IAS steadily increases over the years. In addition,there is substantial variation in the frequency of voluntary IAS adoptionsacross countries.

We begin our tests by analyzing whether voluntary IAS reporting, on av-erage, is associated with higher market liquidity and a lower cost of capitalrelative to local GAAP firms, and relative to firms’ own pre-IAS histories. Weexamine three measures of economic outcomes, namely the price impactof trades (Amihud [2002]), the percentage bid-ask spread, and the impliedcost of capital. Using these variables (and several others in the sensitivityanalyses), we find little evidence that voluntary IAS (or U.S. GAAP) report-ing has beneficial capital-market effects. To benchmark the findings, weshow that proxies indicating stronger firm-level reporting incentives (mea-sured in levels) as well as U.S. cross-listings, which have been shown to rep-resent a credible commitment to more transparency (e.g., Hail and Leuz


[2009]), are associated with higher market liquidity and a lower cost ofcapital. Both benchmark findings make it unlikely that measurement errorin the dependent variables or lack of power are responsible for the weakaverage effects around voluntary IAS adoptions.

Next, we analyze firm-level heterogeneity in the capital-market effectsaround voluntary IAS adoptions. The idea behind these tests is to illus-trate that markets respond positively to changes in firms’ economics thatlead to improvements in management’s reporting incentives around IASadoptions, but not to all IAS adoptions. This, in turn, calls into questionwhether the capital market effects are attributable to IAS reporting perse. We use three proxies to identify major changes in firm-level report-ing incentives around IAS adoptions. Our first proxy is input-based andfocuses directly on firm characteristics that shape management’s reportingincentives. Specifically, we expect managers in firms that are larger, moreprofitable, more international, have larger financing needs, larger growthopportunities, and more dispersed ownership structures to have strongerincentives for transparent financial reporting. We use factor analysis to sum-marize the incentive effects from these firm attributes and extract a singlefactor (with consistent loadings). Our second proxy is output-based and fo-cuses on firms’ reporting behavior. It is built on the idea that actual report-ing changes are ultimately driven by changes in the underlying incentives.Following Leuz, Nanda, and Wysocki [2003], we use the magnitude of ac-cruals relative to the cash flow from operations as a simple characterizationof reporting behavior and earnings informativeness. Accrual-based proxiesfor earnings quality have been contentious in the literature (e.g., Dechow,Ge, and Schrand [2010]). They could also be spuriously affected in our set-ting due to the switch in accounting standards. Thus, we use a third proxythat does not rely on accounting information. Namely, we capture changesin the external reporting environment using the number of analysts fol-lowing a firm. The idea is that scrutiny by analysts and markets also shapesmanagement’s reporting incentives. For each of the three proxies, we com-pute the changes around IAS adoptions, and use the distribution of thechanges to split the sample into “serious” and “label” firms.2

The cross-sectional analyses provide the following results: First, we showthat firms exhibit substantial (positive and negative) variation in the threeproxies for changes in reporting incentives around IAS adoptions. Second,we show that “serious” adopters (i.e., firms with above median changes

2 The construction of our split variables and the focus on changes allows for the possibilitythat firms with strong reporting incentives already provide high-quality reports under localGAAP. Thus, “label” adopters are not necessarily firms with poor reporting, but firms withouta substantial increase in reporting incentives around IAS adoption, and hence one would notexpect to observe positive market effects. For this reason, we control for the level of reportingincentives in the model. Conversely, the “label” group can comprise firms with a substantialreduction in reporting incentives around IAS adoption, thereby leading to negative marketoutcomes.


in reporting incentives) experience an increase in market liquidity and adecline in the cost of capital relative to “label” adopters. The results are sta-tistically and economically significant and similar across partitioning vari-ables. Compared to local GAAP firms, the net effect on market liquidity isalso positive and economically significant, consistent with the notion thatserious adopters increase their commitment to transparency. The net effecton cost of capital is less robust and generally close to zero, except when wesplit by the reporting incentives factor, in which case we find a significantdecrease in the cost of capital. Similarly, we find evidence that the net effecton Tobin’s q is significantly positive, as predicted, for serious adopters whenusing the change in the reporting incentives factor.

Finally, we show that the three partitions also explain firm-level hetero-geneity in the effects around mandatory IFRS adoptions. Thus, the report-ing incentive results carry over to mandatory accounting changes. Thesetests address an important concern: given that our sample spreads overmany years during which IAS/IFRS have evolved substantially, the resultscould reflect heterogeneity in what constitutes an IAS adoption or in thestandards themselves, rather than heterogeneity in the change in report-ing incentives at the time of IAS adoption. Illustrating that the three par-titions produce similar results around mandatory IFRS eases this concernand makes it unlikely that heterogeneity in the IAS coding or in the stan-dards is the primary driver of our results because at the time of the mandateheterogeneity in the standards should be minimal.3 Instead, it points to re-porting incentives as the key source of the heterogeneity in outcomes. Wealso show that serious and label adoptions occur in all time periods andmany countries, and that our main results hold in the early and late part ofthe sample period.

The paper’s two main contributions to the literature are as follows. First,our study is among the first to highlight and test firm-level heterogene-ity in the capital-market effects around voluntary and mandatory IAS/IFRSadoptions. Existing studies tend to focus on the average effect of these adop-tions with respect to some outcome variable or examine cross-sectional dif-ferences at the country level (e.g., Daske et al. [2008], Armstrong et al.[2010], Landsman, Maydew, and Thornock [2012]).4 Our study shows thatfirm-level heterogeneity in reporting incentives plays a significant role forthe capital-market effects around IAS/IFRS adoptions. In doing so, wecontribute to the literature on the importance of firms’ reporting incen-tives (e.g., Ball, Robin, and Wu [2003], Leuz, Nanda, and Wysocki [2003],

3 We further show that the results hold for alternative IAS coding schemes, including a verystrict scheme that requires an explicit auditor’s attestation of IAS/IFRS reporting, as well asfor voluntary U.S. GAAP adoptions.

4 There exist a few studies that provide cross-sectional analyses at the firm level aroundIAS/IFRS adoptions (e.g., Christensen, Lee, and Walker [2007], Karamanou and Nishiotis[2009], Byard, Li, and Yu [2011]). However, they either have much narrower samples or donot focus on the heterogeneity in the effects and the underlying reasons for it.


Burgstahler, Hail, and Leuz [2006]). Much of this literature has focusedon differences in countries’ institutions as a driver of reporting incentives.We extend this literature by applying the notion of reporting incentivesto within-country tests based on firm-level changes, rather than broad andrelatively stable cross-country differences (in levels). Second, our evidencedemonstrates that concerns about selection effects around IAS/IFRS adop-tions are to be taken seriously, particularly for voluntary adoptions, andthat one has to exercise caution in attributing observed capital-market ef-fects solely to the switch in standards. Instead, the effects could also reflectchanges in the economics of the firm that affect managers’ reporting incen-tives. On a more descriptive level, the paper makes a third contribution.It provides comprehensive evidence on voluntary IAS reporting aroundthe globe and highlights that, because of the large number of inconsis-tencies, commonly used accounting standards classifications in Worldscopeand Global Vantage have to be used judiciously.

Section 2 develops our hypotheses. In section 3, we delineate our re-search design and describe the data. Section 4 presents the analyses andresults. Section 5 concludes. In the appendix we provide a comparison ofcommonly used accounting standards classifications and describe voluntaryIAS/IFRS adoptions around the world.

2. Hypothesis Development

The starting point of our study is the notion that the underlying moti-vations for why managers change standards, including changes in firms’economics, play a significant role for the economic consequences aroundIAS/IFRS adoptions. Some firms may adopt IAS merely in name withoutmaking material reporting changes. For others, economic changes (e.g.,new growth opportunities) may provide management with incentives toenhance a firm’s reporting strategy and to strengthen its commitment totransparency. In this case, IAS adoption could be part of a broader set ofchanges. Similarly, there can be heterogeneous incentives around a man-date to report under IFRS; not all firms are likely to equally embrace sucha mandate. As a result of such differences in incentives, and provided thatmarkets are able to differentiate between them, there should be predictableheterogeneity in the economic effects around voluntary and mandatoryIAS/IFRS adoptions.

Important antecedents for this paper are studies highlighting the role offirms’ reporting incentives in explaining observed accounting propertiesand actual practices.5 IAS or IFRS, like any other set of accounting stan-dards, afford management with substantial discretion as their application

5 This literature essentially goes back to Watts and Zimmerman [1986]. Recent examples inthe international literature are Ball, Kothari, and Robin [2000], Ball, Robin, and Wu [2003],Leuz [2003], Burgstahler, Hail, and Leuz [2006], and Bradshaw and Miller [2008]. Thesestudies primarily focus on country-level incentives.


involves judgment and the underlying measurements are often based onprivate information. The way in which firms use this discretion likely de-pends on managers’ reporting incentives, which are shaped by many fac-tors, including countries’ institutional frameworks, various market forces,and firm characteristics. In this paper, we emphasize and test the role offirm-level incentives in the context of IAS/IFRS adoptions.

Our main point is that, because firms differ with respect to their report-ing incentives as well as changes in them over time, it is difficult to attributecapital-market effects around IAS adoptions to the change in the standardsalone. Observed capital-market effects around IAS adoptions likely also re-flect changes in firm economics including the circumstances that led to IASadoption in the first place. This creates a difficult identification and selec-tion issue. To illustrate the point, our main hypothesis is that markets reactmore favorably to IAS adoptions from firms that make material and cred-ible improvements to their overall reporting and disclosure policies. Forinstance, in a firm with a positive shock to its growth opportunities, manage-ment anticipates larger future financing needs, and hence sees more valuein improving transparency.6 If management increases its commitment totransparency and among several changes voluntarily adopts IAS, then theswitch in standards is associated with the change in commitment, but it isnot the sole source of the improvement. It merely serves as a proxy. In con-trast, markets should not react (or could even react negatively) when firmsonly change the label of their accounting standards without making funda-mental changes to their reporting. This heterogeneity provides the motiva-tion for our stylized distinction between “serious” and “label” adopters.

Importantly, we do not claim that “serious” adopters experience positiveeffects because they comply strictly with IAS. We classify them as such be-cause our proxies indicate a substantial change in the underlying incentivesto improve transparency overall. Similarly, we cannot attribute any negativeeffects around “label” adoptions to a lack of compliance with IAS. Instead,the effects could simply reflect reduced incentives for transparent report-ing. Put differently, we do not attempt to identify the marginal effect of IASadoption per se. To the contrary, a key point of our study is to illustratethat the estimated IAS coefficient cannot be attributed to the accountingstandards alone, but likely also reflects differences in firms’ underlying mo-tivations for IAS adoption. Thus, addressing the self-selection inherent inthe voluntary adoption of IAS is not appropriate for our purposes, as we aimto highlight its existence, but doing so would be critical for studies aimingto isolate the effects of IAS reporting. Finding heterogeneity in the capital-market effects around IAS adoption does also not imply that standards donot matter. Standards and, specifically, voluntary IAS adoption can play an

6 This argument is similar to the bonding hypothesis. U.S. cross-listings are predicted to bemore attractive for firms with large financing needs and growth opportunities and to entailhigher costs for firms in which insiders consume substantial private benefits of control (e.g.,Doidge, Karolyi, and Stulz [2004]).


important role in firms’ commitment strategies (e.g., Leuz and Verrecchia[2000]). But the heterogeneity implies that the change in standards aloneis likely not responsible for the observed effects.

In principle, there are several additional explanations for heterogeneouscapital-market effects around IAS adoptions, which we attempt to rule outor address in our research design. First, it is possible that the use of IASharmonizes reporting quality across firms. In this case, the heterogeneityin economic consequences around adoption mainly stems from prior re-porting differences, rather than differences in firms’ motivations for IASadoption. We therefore should see positive capital-market effects for mostIAS adopting firms, considering that IAS are viewed as more demandingand capital-market oriented relative to most local GAAP regimes, and thebiggest effects should occur for the firms with the lowest reporting qualityprior to the switch (all else equal). Given this alternative explanation, wefocus on changes in firms’ reporting incentives around IAS adoption andinclude the pre-IAS level of the reporting proxies in the empirical model tocontrol for prior differences in firms’ reporting practices.

Second, IAS adoption itself could be almost costless, in particular incountries with low-quality institutions (Ball [2006]). One concern then isthat discretion in reporting standards and lack of enforcement make it dif-ficult and very costly for investors to figure out the extent to which firmsimplement serious changes around IAS adoption. Thus, it is conceivablethat there is little (economically meaningful) heterogeneity in the capital-market effects around IAS adoptions because investors cannot discern be-tween serious and label adoptions.

Third, it is possible that the heterogeneity in outcomes around volun-tary IAS adoptions reflects heterogeneity in the standards themselves or inhow firms use the standards. For instance, the set of IAS has changed andevolved considerably over the years (e.g., as a result of the IASC’s Core Stan-dards Project or the Norwalk Agreement). Similarly, firms could use IAS toa varying degree (e.g., some claim to do IAS within the discretion providedby local GAAP while others provide a full-fledged set of IAS statements in-cluding auditor attestation). These differences across time and firms couldlead to differential market reactions regardless of the underlying reportingincentives. To rule out this explanation, we provide additional tests limitingthe heterogeneity in the standards (see section 4.2).

Finally, we note that it is difficult to predict market reactions around “la-bel” adoptions. One prediction is that label adoptions have no or a negli-gible effect. Alternatively, the market reaction could be adverse, as a labeladoption makes it apparent to investors that a firm is unwilling to com-mit to more transparency. Moreover, label adoptions could increase in-vestor uncertainty, for example, because the change in standards makesit harder to forecast future earnings. But if markets unravel label adop-tions and react unfavorably to them, it poses the question of why firmschoose such adoptions in the first place. Again, it is important to recognizethat the capital-market effects around “label” adoptions are not necessarily


attributable to the switch in standards per se. A negative market reactioncould simply reflect changes in firm economics including a correspondingdecrease in the commitment to transparency, rather than a response to theadoption of IAS. It is also possible that label adoptions are not very costly,yet offer (small) benefits that we do not measure in our analysis. For in-stance, firms could adopt IAS for contracting reasons. IAS could also be aninvestor-relations tool to improve communications with foreign investors,which could have an impact on foreign holdings, but less of an effect onliquidity (as one group of investors simply replaces another).7 It could alsobe that managers perceive a general trend to IAS and feel compelled to fol-low, but adopt IAS in the least costly way because there is no reason to ad-just the reporting practices. Given all these possibilities, we do not sign ourexpectation for the effects of label adoptions relative to local GAAP firms.8

3. Research Design and Data

Our research design relies on variation in the underlying motives forIAS/IFRS adoption. We therefore focus on voluntary adoptions in our mainanalyses to increase the power of the tests, and use changes in firms’ report-ing incentives to illustrate the selection issue. Doing so, we need a variableindicating when a firm has voluntarily adopted IAS, a variable capturingthe changes in firms’ reporting incentives to differentiate between “seri-ous” and “label” adopters, as well as proxies for the economic outcomesand a set of controls. We estimate the following model:

EconCon = β0 + β1IAS + β2Serious Adopters +∑

β j Controls j + ε. (1)

EconCon stands for three proxies of economic consequences (i.e., price im-pact, bid-ask spreads, and cost of capital). IAS is a binary variable coded as“1” for years in which a firm follows IAS. Serious Adopters denotes a binaryclassification that identifies firms with above-median changes in our report-ing proxies around the adoption of IAS. With this model, we can comparethe label adopters to the local GAAP firms (β1), the serious adopters to thelabel adopters (β2), and the serious adopters to the base group (β1 + β2).Controlsj denotes a vector of control variables including fixed-effects. Thenext sections describe each of the model’s components.

3.1 IAS REPORTING AND SERIOUS VERSUS LABEL CLASSIFICATION

The coding of the IAS variables in equation (1) involves three steps. First,we construct a firm-year panel data set with a binary IAS reporting variable.

7 For instance, Tan, Wang, and Welker [2011] show that the IFRS mandate has differentialeffects on analyst forecasts depending on whether the analysts are local or foreign.

8 Yet another explanation is that managers do not believe that markets unravel label adop-tions. Prior capital-market studies suggest that managers sometimes engage in accountingchoices as if markets reward or fixate on these choices, despite the fact that studies showthat markets see through these choices (e.g., Watts and Zimmerman [1986], Kothari [2001]).


We deliberately use a broad classification when coding the binary IAS vari-able. Given our research question, we intend to capture a wide variety ofadoption strategies, including firms with dual reporting, those that providea reconciliation of local GAAP numbers to IAS, and those that merely createthe appearance of IAS reporting. That is, our classification relies primarilyon what firms claim they do. Second, we determine the actual switch year,i.e., the point in time when IAS reporting started. We provide details onthese two steps in the appendix.

As a third step, we construct three proxies measuring changes in report-ing incentives around the switch to IAS. We use these proxies to partitionthe IAS firms into serious and label adopters. The underlying idea is thatreporting incentives play a significant role for firms’ reporting strategiesas well as the choice to voluntarily adopt IAS. Based on this logic, we at-tempt to identify firms that experience substantial increases in their report-ing incentives around IAS adoption. These firms are likely to make majorimprovements to their reporting strategy. In contrast, there could be firmsthat adopt IAS without material changes or even decreases in reporting in-centives. For example, these firms could adopt IAS because managers per-ceive a general trend to IAS/IFRS and feel pressured to follow, but then areless likely to materially adjust their reporting. Similarly, they could adoptIAS for contracting reasons and hence not be concerned with transparencyto equity investors.

Since reporting incentives are unobservable, we create three distinctproxies for the underlying construct: two focus on the determinants offirms’ incentives (and hence are input-based); one relies on firms’ actualreporting behavior as a result of firms’ incentives (and hence is output-based). Our first partitioning variable reflects observable firm character-istics. Economic theory suggests that larger, more profitable firms withgreater financing needs and growth opportunities, more international op-erations, and dispersed ownership have stronger incentives for transparentreporting to outside investors.9 We measure the Reporting Incentives variableas the first and primary factor (out of three that are retained) when apply-ing factor analysis to the following six firm attributes: firm size (natural logof the US$ market value), financial leverage (total liabilities over total as-sets), profitability (return on assets), growth opportunities (book-to-marketratio), ownership concentration (percentage of closely held shares), and in-ternationalization (foreign sales over total sales). The factor exhibits all theexpected loadings (i.e., increasing in size, leverage, profitability, growth,and foreign sales; decreasing in ownership concentration).10

9 These determinants follow from the disclosure literature (see Leuz and Wysocki [2008]for a survey) and the cross-listing literature (e.g., Lang, Lins, and Miller [2003], Doidge,Karolyi, and Stulz [2004]).

10 To maximize sample size we replace missing values for ownership concentration andforeign sales with zero. Unreported sensitivity analyses show that our results do not hinge onthe composition of the factor score. When we rerun our analyses eliminating, one-by-one, each


Our second partitioning variable relies on a simple accrual-based charac-terization of actual reporting. Sloan [1996] or Bradshaw, Richardson, andSloan [2001] show that the decomposition of earnings into accruals andoperating cash flow as well as extreme accruals contain important informa-tion. Furthermore, the ratio of accruals to cash flows has been shown to pro-duce plausible earnings management rankings for firms around the world(Leuz, Nanda, and Wysocki [2003], Wysocki [2004]).11 Following Leuz,Nanda, and Wysocki [2003], we compute the Reporting Behavior variableas the ratio of the absolute value of accruals to the absolute value of cashflows (multiplied by –1 so that higher values indicate more transparent re-porting). Scaling by the operating cash flow serves as a performance adjust-ment (Kothari, Leone, and Wasley [2005]). We estimate accruals as the dif-ference between net income before extraordinary items and the cash flowfrom operations or, if unavailable, compute them following the balance-sheet approach in Dechow, Sloan, and Sweeney [1995].12

The idea behind the third partitioning variable is to capture externalchanges affecting firms’ reporting incentives, such as the scrutiny by ana-lysts and financial markets. Lang and Lundholm [1996] show that analystcoverage is related to more transparent reporting. Evidence in Lang, Lins,and Miller [2004] and Yu [2008] suggests a monitoring role of financialanalysts, for instance, in curbing earnings management. We compute theReporting Environment variable as the natural log of the number of analystsfollowing the firm (plus one). For firms without coverage in I/B/E/S weset analyst following to zero. A nice feature of this variable is that it doesnot rely on accounting information, and is free of mechanical accountingeffects that likely occur around the switch to a new set of accountingstandards.13

firm attribute (or replacing market value with total assets and dropping the book-to-marketratio), the results are not materially affected.

11 We acknowledge that it is difficult to consistently assess the actual reporting behavioracross many firms and countries. To mitigate these concerns, we (1) use the reporting variablesonly to partition the IAS firms and not directly as test variables, (2) base all our cross-sectionalanalyses on changes in the reporting incentives proxies, (3) do not analyze changes in theaccruals proxy per se but focus on the capital-market consequences of IAS adoption, and (4)always include the (pre-IAS) level of our partitioning variable as a control.

12 We recognize that a change in accounting standards could have mechanical effects onthe magnitude of accruals. However, as IAS tend to be more accruals-based than local stan-dards around the world (e.g., Hung [2001], Ding et al. [2006]), this effect likely works againstour expectations. In addition, we control for the variability in firms’ operations as suggestedby Hribar and Nichols [2007]. That is, we first regress the ratio of absolute accruals to cashflows on the standard deviation of cash flows, and then use the residuals as our Reporting Be-havior variable. This procedure accounts for the variability in operations that potentially biasesunsigned accruals measures. The results (not tabulated) remain largely unchanged using thisalternative metric.

13 We find very similar results using analyst forecast dispersion as a proxy for firms’ ReportingEnvironment (not tabulated).


Next, to reduce measurement errors and allow for the possibility thatincentives change slowly over time, we compute the rolling average (overthe years t, t − 1, t − 2, relative to the year t of IAS adoption) for eachreporting variable. We then use this rolling average in two ways: first, weinclude it as a separate control in equation (1) for differences in the levelof firms’ reporting incentives. Higher values indicate stronger incentivesfor transparent reporting. Second, we compute changes in the reportingvariables around IAS adoption by subtracting the rolling average in yeart − 1 (computed over the years t − 3 to t − 1) from the rolling averagein year t + 3 (computed over the years t + 1 to t + 3). We center the com-putation on year t = 0 to obtain nonoverlapping observations from the pre-and postperiod and to exclude potentially contaminating effects from theadoption year itself. We then use the distribution of the changes aroundIAS adoption (with each firm represented once) to classify IAS firms withabove median changes as Serious Adopters (coded as “1”) and with belowmedian changes as label adopters (see also panel A in table 4).14

We assess the construct validity of the three reporting variables by com-puting the Spearman correlations between the proxies (ranging from ρ =0.18 between Reporting Environment and Reporting Behavior to ρ = 0.54 be-tween Reporting Environment and Reporting Incentives), and by correlatingthem with related reporting variables, for which we have only limited sam-ples at hand. For instance, we find significantly positive correlations forthe Reporting Incentives and the Reporting Environment variables with (1) adisclosure quality index created by business professionals in Austria, Ger-many, and Switzerland (Daske and Gebhardt [2006]), (2) having a BigFive auditor, and (3) the extent of disclosure (measured by the number ofpages in the annual report, after adjusting for country effects, firm size, andperformance).15

3.2 DEPENDENT VARIABLES

In studying the economic consequences around IAS adoptions, we useproxies for market liquidity, information asymmetry, and the cost of cap-ital, which—among other things—should reflect the quality of financialdisclosure and reporting (e.g., Verrecchia [2001], Lambert, Leuz, andVerrecchia [2007]). This approach sidesteps the difficulties of explicitlymeasuring changes in reporting quality around IAS adoption.

The first dependent variable is a measure of illiquidity suggested by Ami-hud [2002] and inspired by Kyle’s [1985] lambda. We measure Price Impact,

14 In untabulated sensitivity analyses, we also compute the cutoff value based on the entiredistribution of changes in the respective reporting variable around any given year (thus in-cluding non-IAS firms and years other than the IAS adoption year). Alternatively, we rank thechanges in the reporting variables around IAS adoption into deciles, and then use the ranks toestimate the differential economic consequences of serious and label adopters. In each case,the results look similar to those in the main analysis and none of the inferences change.

15 Ultimately, partitions based on the three proxies are unlikely to produce consistent andsensible results if they fail to capture the underlying construct. In this sense, the cross-sectionalanalyses can also be seen as a validation exercise of the three partitioning variables.


i.e., the ability of an investor to trade in a stock without moving its price,as the yearly median of the Amihud [2002] illiquidity measure (computeddaily and equal to the absolute stock return divided by the US$ trading vol-ume). Higher values indicate more illiquid stocks. To avoid the misclassifi-cation of days with no or low trading activity (i.e., days potentially yieldinga price impact of zero), we omit zero-return days from the computation ofthe yearly medians.

The second dependent variable is the Bid-Ask Spread, which is a com-monly used proxy for information asymmetry (e.g., Welker [1995], Leuzand Verrecchia [2000], Lang, Lins, and Maffett [2012]). We use the yearlymedian of the daily quoted spreads, which we compute as the differencebetween the closing bid and ask prices divided by the midpoint.

The third dependent variable is Cost of Capital . We follow Hail andLeuz [2006] and compute the mean implied cost of equity capital usingfour models suggested by Claus and Thomas [2001], Gebhardt, Lee, andSwaminathan [2001], Easton [2004], and Ohlson and Juettner-Nauroth[2005]. These models are consistent with discounted dividend valuationbut rely on different earnings-based representations. We substitute marketprice and analyst forecasts into each valuation equation and back out thecost of capital as the internal rate of return that equates current stock priceand the expected future sequence of residual incomes or abnormal earn-ings. See the appendix in Hail and Leuz [2006] for details on the cost ofcapital proxies we apply.16

We measure the dependent variables as of month +10 after the fiscal yearend for which we code the accounting standards. We choose this month toensure that firms’ annual reports are publicly available and priced at thetime of the computations. For variables that are computed over an entireyear, we start the computation as of month –2 through month +10 relativeto the firm’s fiscal year end.

3.3 CONTROL VARIABLES

We include industry-, country-, and year-fixed effects in all regressionmodels, and hence control for differences in countries’ adoption ratesas well as time trends. In unreported analyses, we also check that our re-sults are robust when we include country-year-fixed effects to control forcountry-wide shifts in the adoption rates over time, e.g., due to the an-nouncement of mandatory IFRS reporting. Thus, the Serious Adopters coef-ficient reflects within-country differences relative to local GAAP firms andlabel adopters.

16 We recognize that there is a debate about the empirical validity of implied cost of capitalestimates (e.g., Botosan and Plumlee [2005], Easton and Monahan [2005]). One alternativeis to use realized returns as a proxy for expected returns. However, this proxy has many draw-backs as well, especially with short time series (Elton [1999]). We therefore go down a differentroute and use proxies for liquidity and information asymmetry as these constructs also capturedifferences in reporting quality (Leuz [2003], Lang, Lins, and Maffett [2012]).


In addition, we introduce the following binary indicators (see also Daskeet al. [2008]): U.S. Listing is equal to one if the shares are traded over thecounter or cross-listed on a U.S. exchange (Hail and Leuz [2009]). We sep-arately code firms that voluntarily report under U.S. GAAP , but without aU.S. cross-listing. We set New Markets equal to one if shares are traded on anexchange that specializes in technology and other high-growth stocks andhas listing requirements that mandate or allow IAS reporting. Index Memberrepresents firms whose shares are constituents of national or internationalstock market indices as defined in Worldscope.

In the liquidity regressions, we control for firm size, share turnover, andreturn variability (Chordia, Roll, and Subrahmanyam [2000], Leuz andVerrecchia [2000]). Firm size is the Market Value of equity in US$ million.Share Turnover is the accumulated US$ trading volume divided by the mar-ket value of outstanding equity. Return Variability is computed as annualstandard deviation of monthly stock returns. We compute share turnoverand return variability from month –2 through month +10 relative to afirm’s fiscal year end, and lag the market variables by one year to mitigateconfounding effects from contemporaneous measurement.

In the cost of capital specifications, we follow Hail and Leuz [2009] andcontrol for firm size measured as Total Assets (in US$ million), FinancialLeverage equal to the ratio of total liabilities to total assets, and Return Vari-ability. We also control for Forecast Bias equal to the one-year-ahead analystforecast error scaled by lagged total assets. First, it is possible that the adop-tion of IAS impairs analysts’ ability to forecast earnings, at least during atransitional period. Second, any bias in analyst forecasts could mechanicallyaffect our implied cost of capital estimate if markets back out the bias (e.g.,Botosan and Plumlee [2005], Hail and Leuz [2006]). Finally, we controlfor inflation because analyst forecasts are expressed in nominal terms andlocal currency. We measure Inflation as the yearly median of one-year-aheadrealized monthly changes in the consumer price index in a country.17

3.4 DATA

We obtain financial data from Worldscope; return, bid-ask spread, andtrading volume data from Datastream; and analyst forecasts and share pricedata for the cost of capital estimation from I/B/E/S. The sample consistsof all Worldscope firms from 1990 to 2005 for which we have the necessarydata to compute the variables described above.18 Table 1 presents the dis-tribution of firms reporting under IAS/IFRS, U.S. GAAP, or local GAAP bycountry (panel A) and year (panel B). The total sample consists of 69,528

17 Using countries’ risk-free rates rather than the inflation rate yields very similar resultsand inferences.

18 Our main sample ends on December 31, 2005, because IFRS became mandatory in manycountries thereafter. We further require firms to have total assets of 10 US$ million, and coun-tries to have at least one voluntary IAS observation. The latter two design choices do not affectthe inferences drawn from the analyses.


TA

BL

E1

Sam

ple

Com

posi

tion

byC

ount

ryan

dYe

ar

Pan

elA

:Acc

ount

ing

Stan

dard

s,L

isti

ngSt

atus

,and

Inde

xM

embe

rshi

pby

Cou

ntry

IAS/

IFR

SU

.S.G

AA

PU

.S.L

isti

ng

New

Mar

kets

Inde

xM

embe

r

Cou

ntr

yU

niq

ueFi

rms

Firm

-Yea

rsFi

rm-Y

ears

%Fi

rm-Y

ears

%Fi

rm-Y

ears

%Fi

rm-Y

ears

%Fi

rm-Y

ears

%A

ustr

alia

983

5,41

266

1.2

30.

170

313

.058

1.1

2,89

253

.4A

ustr

ia13

197

422

322

.920

2.1

939.

546

4.7

221

22.7

Bel

gium

171

1,15

978

6.7

80.

727

2.3

00.

030

226

.1B

erm

uda

1866

1218

.225

37.9

913

.612

18.2

1218

.2C

anad

a1,

277

7,55

47

0.1

122

1.6

1,45

819

.335

0.5

2,80

237

.1C

hin

a1,

121

3,36

272

821

.721

0.6

127

3.8

60.

289

2.6

Cze

chR

epub

lic55

159

4226

.40

0.0

85.

00

0.0

2113

.2D

enm

ark

239

2,12

990

4.2

20.

116

0.8

00.

030

914

.5Fi

nla

nd

152

919

283.

00

0.0

242.

60

0.0

226

24.6

Fran

ce98

56,

253

132

2.1

440.

719

73.

20

0.0

2,49

940

.0G

erm

any

717

2,97

275

925

.536

312

.210

13.

473

724

.856

719

.1G

reec

e33

12,

131

261.

24

0.2

361.

70

0.0

616

28.9

Hon

gK

ong

730

4,87

160

1.2

130.

361

312

.650

1.0

1,18

624

.3H

unga

ry35

207

125

60.4

62.

931

15.0

00.

012

5.8

Isra

el15

262

918

2.9

158

25.1

6911

.026

4.1

190

30.2

Ital

y18

81,

024

376

36.7

00 .

026

2.5

272.

620

720

.2L

uxem

bour

g29

142

3222

.517

12.0

2316

.24

2.8

6747

.2T

he

Net

her

lan

ds28

82,

187

442.

073

3.3

167

7.6

411.

928

312

.9N

orw

ay24

01,

619

150.

912

0.7

122

7.5

00.

036

122

.3Pa

kist

an86

583

447.

50

0.0

00.

00

0.0

449

77.0

Peru

5428

274

26.2

41.

438

13.5

00.

098

34.8

Pola

nd

103

519

295.

62

0.4

183.

50

0.0

9718

.7Po

rtug

al10

059

910

1.7

00.

024

4.0

00.

027

345

.6R

ussi

anFe

dera

tion

3615

348

31.4

149.

287

56.9

00.

06

3.9

Sin

gapo

re45

62 ,

367

251.

112

0.5

109

4.6

1,87

979

.472

930

.8So

uth

Afr

ica

494

2,76

288

3.2

10.

037

013

.40

0.0

851

30.8

Swed

en37

02,

408

401.

70

0.0

157

6.5

70.

337

515

.6Sw

itze

rlan

d28

32,

076

747

36.0

301.

434

1.6

80.

428

713

.8Tu

rkey

193

1,10

018

016

.40

0.0

252.

30

0.0

656

59.6

Un

ited

Kin

gdom

2,15

412

,910

90.

113

0.1

1,14

48.

974

75.

85,

922

45.9

Tota

l12

,171

69,5

284,

155

6.0

967

1.4

5,85

68.

43,

683

5.3

22,6

0532

.5(C

ontin

ued)


TA

BL

E1—

Con

tinue

d

Pan

elB

:Acc

ount

ing

Stan

dard

s,L

isti

ngSt

atus

,and

Inde

xM

embe

rshi

pby

Year

IAS/

IFR

SU

.S.G

AA

PU

.S.L

isti

ng

New

Mar

kets

Inde

xM

embe

r

Year

Firm

sFi

rms

%Fi

rms

%Fi

rms

%Fi

rms

%Fi

rms

%19

901,

225

262.

14

0.3

112

9.1

383.

147

638

.919

911,

877

432.

38

0.4

154

8.2

432.

370

037

.319

922,

371

592.

59

0.4

179

7.5

622.

683

835

.319

932,

612

612.

311

0.4

189

7.2

612.

388

633

.919

942,

983

923.

115

0.5

226

7.6

551.

898

232

.919

953,

252

101

3.1

170.

525

98.

089

2.7

1,08

033

.219

963,

633

139

3.8

230.

632

38.

913

33.

71,

214

33.4

1997

4,35

519

34.

425

0.6

396

9.1

157

3.6

1,40

132

.219

984,

664

220

4.7

310.

744

59.

516

93.

61,

569

33.6

1999

5,49

631

85.

859

1.1

503

9.2

232

4.2

1,83

733

.420

005,

904

429

7.3

941.

653

49.

034

85.

91,

983

33.6

2001

6,85

355

28.

116

32.

462

29.

157

98.

42,

212

32.3

2002

7,22

461

68.

516

12.

263

78.

866

09.

12,

342

32.4

2003

7,65

061

88.

115

72.

158

17.

649

46.

52,

215

29.0

2004

7,32

163

38.

616

32.

253

07.

238

15.

22,

082

28.4

2005

2,10

855

2.6

271.

316

67.

918

28.

678

837

.4To

talF

irm

-Yea

rs69

,528

4,15

56.

096

71.

45,

856

8.4

3,68

35.

322

,605

32.5

Th

esa

mpl

eco

mpr

ises

am

axim

umof

69,5

28fi

rm-y

ear

obse

rvat

ion

sfr

om30

coun

trie

sw

ith

fisc

alye

aren

dsbe

twee

nJa

nua

ry1,

1990

,an

dD

ecem

ber

31,2

005,

for

wh

ich

we

hav

esu

ffici

ent

Wor

ldsc

ope

and

Dat

astr

eam

data

toes

tim

ate

our

base

regr

essi

ons

(see

tabl

e3)

.We

excl

ude

fisc

alye

ars

offi

rms

that

are

subj

ect

tom

anda

tory

IFR

Sre

port

ing.

We

requ

ire

firm

sto

hav

eto

tala

sset

sof

10U

S$m

illio

nor

mor

e,an

dlim

itth

esa

mpl

eto

coun

trie

sw

ith

atle

asto

ne

volu

nta

ryIA

Sfi

rm-y

ear

obse

rvat

ion

.Th

eta

ble

repo

rts

the

num

ber

ofun

ique

firm

sas

wel

las

firm

-yea

rob

serv

atio

ns

and

perc

enta

ges

byco

untr

y(p

anel

A)

and

year

(pan

elB

)fo

rth

efo

llow

ing

indi

cato

rva

riab

les

(cod

edas

“1”

ifth

ede

fin

itio

nap

plie

s):I

AS/

IFR

San

dU

.S.G

AA

Pin

dica

tefi

nan

cial

repo

rtsi

nac

cord

ance

wit

hth

ere

spec

tive

acco

unti

ng

stan

dard

s.W

eid

enti

fyfi

rms’

repo

rtin

gst

anda

rdsb

ased

onth

e“a

ccou

nti

ng

stan

dard

sfol

low

ed”

fiel

din

Wor

ldsc

ope

(fiel

d07

536)

,an

dad

just

the

codi

ng

for

con

trad

icto

ryin

form

atio

nfr

oman

exte

nsi

vem

anua

lrev

iew

offi

rms’

ann

ualr

epor

ts(s

eeth

eap

pen

dix

for

deta

ils).

U.S

.L

istin

gm

arks

firm

-yea

rsfr

omco

mpa

nie

sw

hos

esh

ares

are

trad

edov

er-th

e-co

unte

ror

are

exch

ange

-list

edin

the

Un

ited

Stat

es(s

eeH

ail

and

Leu

z[2

009]

).W

edo

not

incl

ude

the

U.S

.lis

tin

gob

serv

atio

ns

inth

eU

.S.G

AA

Pin

dica

tor.

New

Mar

ket

obse

rvat

ion

sst

emfr

omfi

rms

trad

edon

anex

chan

geth

atsp

ecia

lizes

inte

chn

olog

ysh

ares

and

oth

erh

igh

-gro

wth

stoc

ks,a

ndth

ath

aslis

tin

gre

quir

emen

tsm

anda

tin

gor

allo

win

gfi

nan

cial

repo

rts

inac

cord

ance

wit

hIA

S/IF

RS

(i.e

.,A

lter

nat

ive

Inve

stm

entM

arke

tin

the

Un

ited

Kin

gdom

,Exp

andi

Mar

keti

nIt

aly,

Neu

erM

arke

tin

Ger

man

y,N

ordi

cG

row

thM

arke

tin

Swed

en,a

nd

Sesd

aqin

Sin

gapo

re).

Th

eIn

dex

Mem

ber

vari

able

repr

esen

tsfi

rms

wh

ose

shar

esar

eco

nst

itue

nts

ofn

atio

nal

orin

tern

atio

nal

stoc

km

arke

tin

dice

sas

defi

ned

inW

orld

scop

e(fi

eld

0566

1).


firm-year observations across 30 countries, of which 4,155 are coded asIAS.19 The number of IAS firms increases considerably over time. By 2004,almost 9% of the sample firms have adopted IAS. Similarly, the number ofU.S. GAAP firms increases, but remains at a relatively small 2%.

Table 2 presents descriptive statistics for the dependent variables(panel A) and the control variables (panel B) across the IAS adopters andlocal GAAP firms. Spread and price impact are right-skewed and could ex-hibit multiplicative relations with the control variables. Thus, as is commonin the literature on micro-structure models, we estimate a log-linear speci-fication and use the natural log of these measures in the analyses.

4. Results

4.1 AVERAGE EFFECTS ACROSS IAS AND LOCAL GAAP FIRMS

We begin our analysis by examining average differences in market liquid-ity and cost of capital between firms reporting under IAS and firms report-ing under local GAAP. We use cross-sectional, time-series panel regressions,which benchmark IAS firms against local GAAP firms and against theirown local GAAP history before adoption. Estimating average effects allowsa comparison with prior work. We estimate the empirical specification inequation (1), but without the Serious Adopters variable. For each dependentvariable, we estimate three regressions using either the level of the ReportingIncentives, the Reporting Behavior , or the Reporting Environment variable as acontrol. If these proxies capture incentives for more transparent reporting,they should exhibit a negative coefficient in the model. Table 3 presentsresults from OLS regressions with robust standard errors that are clusteredby firm.20

Using price impact as dependent variable, the coefficients on IAS areinsignificant, suggesting that firms voluntarily reporting under IAS—onaverage—have similar liquidity as local GAAP firms. The control variablesbehave as expected. Importantly, all three reporting variables are negativelyassociated with price impact, consistent with the idea that these proxies cap-ture firms’ incentives for transparent reporting. Moreover, cross-listing inthe United States, which among other things represents a commitment to

19 In several sample countries, IAS reporting was not permissible to satisfy local reportingrequirements during the sample period (e.g., the United Kingdom and Canada). Nonethe-less, firms in these countries could voluntarily provide a second set of financial statementsin accordance with IAS/IFRS while filing primary accounts that comply with local reportingrequirements. We check that our results are not affected by the inclusion of these countries.

20 We also estimate the average effects model (1) replacing the country- and industry-fixedeffects with firm-fixed effects, and (2) omitting the IAS firms without identifiable switch year.The first test essentially controls for time-invariant and potentially unobserved differencesbetween firms. The second test limits the sample to the same set of IAS firms that we use inthe subsequent cross-sectional analyses. The results of both sensitivity tests (not tabulated) aresimilar to those reported in the tables, unless we discuss differences in the text.


TA

BL

E2

Des

crip

tive

Stat

istic

sfo

rR

egre

ssio

nVa

riab

les

Acr

oss

Volu

ntar

yIA

San

dL

ocal

GA

AP

Rep

ortin

gFi

rms

Acc

oun

tin

gVa

riab

leSt

anda

rdN

Mea

nSt

d.D

ev.

P1P2

5M

edia

nP7

5P9

9

Pan

elA

:Dep

ende

ntVa

riab

les

Pric

eIm

pact

Loc

alG

AA

P65

,373

2.64

49.

903

0.00

00.

028

0.19

11.

191

49.4

04IA

S4,

155

1.55

66.

582

0.00

10.

017

0.09

90.

601

26.2

57B

id-A

skSp

read

Loc

alG

AA

P45

,593

0.03

30.

040

0.00

20.

009

0.01

90.

039

0.20

5IA

S3,

450

0.01

90.

024

0.00

10.

006

0.01

20.

024

0.10

5C

osto

fCap

ital

Loc

alG

AA

P24

,134

11.9

%4.

1%5.

5%9.

0%11

.1%

14.0

%25

.1%

IAS

1,80

012

.1%

4.3%

5.7%

8.9%

11.4

%14

.2%

25.8

%

Pan

elB

:Ind

epen

dent

Vari

able

s

Rep

ortin

gIn

cent

ives

Loc

alG

AA

P63

,949

0.03

10.

562

−1.3

26−0

.349

0.04

10.

413

1.25

5IA

S4,

127

0.16

80.

567

−1.2

11−0

.203

0.17

10.

571

1.33

0R

epor

ting

Beh

avio

rL

ocal

GA

AP

57,1

24−1

.637

2.56

2−1

3.39

7−1

.599

−0.8

09−0

.470

−0.0

86IA

S3,

215

−1.6

322.

636

−13.

441

−1.5

01−0

.822

−0.5

15−0

.106

Rep

ortin

gEn

viro

nmen

tL

ocal

GA

AP

65,3

731.

009

1.01

10.

000

0.00

00.

732

1.86

63.

144

IAS

4,15

51.

083

1.07

20.

000

0.00

00.

828

2.04

93.

198

Mar

ketV

alue

Loc

alG

AA

P65

,373

582

1,49

64

3712

241

08,

267

IAS

4,15

584

51,

752

672

238

721

9,34

9Sh

are

Turn

over

Loc

alG

AA

P65

,373

0.55

90.

900

0.00

30.

114

0.31

20.

662

4.44

9IA

S4,

155

0.52

60.

910

0.00

20.

098

0.26

70.

593

4.31

0R

etur

nVa

riab

ility

Loc

alG

AA

P65

,373

0.12

10.

080

0.02

30.

068

0.09

90.

151

0.42

1IA

S4,

155

0.13

30.

083

0.02

10.

075

0.11

00.

168

0.41

7To

talA

sset

sL

ocal

GA

AP

64,8

521,

216

3,79

611

6419

167

519

,954

IAS

4,11

61,

971

4,94

416

120

404

1,40

025

,301

(Con

tinue

d)


TA

BL

E2—

Con

tinue

d

Acc

oun

tin

gVa

riab

leSt

anda

rdN

Mea

nSt

d.D

ev.

P1P2

5M

edia

nP7

5P9

9

Pan

elB

:Ind

epen

dent

Vari

able

s

Fina

ncia

lLev

erag

eL

ocal

GA

AP

64,2

320.

508

0.24

10.

012

0.34

00.

520

0.67

80.

964

IAS

4,09

50.

529

0.23

60.

013

0.36

60.

548

0.70

00.

968

Fore

cast

Bia

sL

ocal

GA

AP

34,9

120.

010

0.04

4−0

.079

−0.0

040.

001

0.01

30.

213

IAS

2,60

60.

011

0.04

5−0

.088

−0.0

040.

001

0.01

50.

214

Infla

tion

–68

,273

2.4%

1.9%

0.0%

1.4%

2.1%

3.0%

9.6%

Th

esa

mpl

eco

mpr

ises

am

axim

umof

69,5

28fi

rm-y

ear

obse

rvat

ion

sfr

om30

coun

trie

sbe

twee

n19

90an

d20

05w

ith

fin

anci

alda

tafr

omW

orld

scop

ean

dpr

ice/

volu

me

data

from

Dat

astr

eam

(see

tabl

e1)

.IA

San

dL

ocal

GA

AP

repr

esen

tfi

rm-y

ears

wit

hfi

nan

cial

repo

rts

inac

cord

ance

wit

hei

ther

IAS/

IFR

Sor

dom

esti

cac

coun

tin

gst

anda

rds,

base

don

our

augm

ente

dW

orld

scop

eac

coun

tin

gst

anda

rds

clas

sifi

cati

onde

scri

bed

inth

eap

pen

dix.

Th

eta

ble

repo

rts

desc

ript

ive

stat

isti

csfo

rth

ede

pen

den

tva

riab

les

(pan

elA

)an

dth

eco

nti

nuo

usin

depe

nde

ntv

aria

bles

(pan

elB

)ac

ross

IAS

and

loca

lGA

AP

firm

-yea

rob

serv

atio

ns.

We

use

thre

ede

pen

den

tvar

iabl

esin

our

prim

ary

anal

yses

:(1)

Pric

eIm

pact

isth

eye

arly

med

ian

ofth

eA

mih

ud[2

002]

illiq

uidi

tym

easu

re(i

.e.,

daily

abso

lute

stoc

kre

turn

divi

ded

byU

S$tr

adin

gvo

lum

e).(

2)T

he

Bid

-Ask

Spre

adis

the

year

lym

edia

nqu

oted

spre

ad(i

.e.,

diff

eren

cebe

twee

nth

ebi

dan

das

kpr

ice

divi

ded

byth

em

id-p

oin

tan

dm

easu

red

atth

een

dof

each

trad

ing

day)

.(3)

Cos

tofC

apita

lis

the

aver

age

cost

ofca

pita

lest

imat

eim

plie

dby

the

mea

nI/

B/E

/San

alys

tcon

sen

sus

fore

cast

san

dst

ock

pric

esus

ing

the

valu

atio

nm

odel

sof

Cla

usan

dT

hom

as[2

001]

,Geb

har

dt,L

eean

dSw

amin

ath

an[2

001]

,Eas

ton

[200

4],

and

Oh

lson

and

Juet

tner

-Nau

roth

[200

5].

See

the

appe

ndi

xin

Hai

lan

dL

euz

[200

6]fo

rde

tails

onth

eim

plie

dco

stof

capi

tal

esti

mat

ion

proc

edur

e.In

each

ofou

rre

gres

sion

mod

els,

we

acco

unt

for

firm

s’re

port

ing

prac

tice

sw

ith

one

ofth

efo

llow

ing

thre

eva

riab

les:

(1)

Usi

ng

fact

oran

alys

is,

we

extr

act

asi

ngl

efa

ctor

indi

cati

ng

the

stre

ngt

hof

firm

s’re

port

ing

ince

nti

ves

from

vari

ous

firm

attr

ibut

es(i

.e.,

mar

ket

valu

eof

equi

ty,fi

nan

cial

leve

rage

,ret

urn

onas

sets

,boo

k-to

-mar

ket

rati

o,pe

rcen

tof

clos

ely

hel

dsh

ares

,an

dpe

rcen

tof

fore

ign

sale

s).W

eth

enco

mpu

teth

ele

vel

ofa

firm

’sR

epor

ting

Ince

ntiv

esin

year

tas

the

rolli

ng

aver

age

ofth

era

wfa

ctor

scor

esov

erth

eye

ars

tto

t−

2.H

igh

erva

lues

den

ote

grea

ter

repo

rtin

gin

cen

tive

s.(2

)W

em

easu

refi

rms’

actu

alre

port

ing

beh

avio

ras

the

abso

lute

valu

eof

accr

uals

scal

edby

the

abso

lute

valu

eof

cash

flow

sfr

omop

erat

ion

s.W

eth

enco

mpu

teth

ele

velo

fafi

rm’s

Rep

ortin

gB

ehav

ior

inye

art

asth

ero

llin

gav

erag

eof

the

raw

mea

sure

sov

erth

eye

ars

tto

t−

2.W

em

ulti

ply

the

met

ric

by–1

soth

ath

igh

erva

lues

den

ote

less

earn

ings

man

agem

ent

and

mor

etr

ansp

aren

tre

port

ing.

(3)

We

mea

sure

exte

rnal

pres

sure

from

the

repo

rtin

gen

viro

nm

ent

byth

en

umbe

rof

anal

ysts

follo

win

gth

efi

rm.F

orfi

rms

wit

hou

tco

vera

gein

I/B

/E/S

we

set

anal

yst

follo

win

gto

zero

.We

then

com

pute

the

leve

lof

afi

rm’s

Rep

ortin

gEn

viro

nmen

tin

year

tas

the

rolli

ng

aver

age

ofth

en

atur

allo

gsof

the

ann

ualn

umbe

rof

anal

ysts

(plu

son

e)ov

erth

eye

ars

tto

t−

2.H

igh

erva

lues

den

ote

mor

eex

tern

alpr

essu

re.T

he

rem

ain

ing

con

trol

vari

able

sar

e:M

arke

tVal

ueis

stoc

kpr

ice

tim

esth

en

umbe

rof

shar

esou

tsta

ndi

ng

(in

US$

mill

ion

).Sh

are

Turn

over

isan

nua

lUS$

trad

ing

volu

me

divi

ded

bym

arke

tva

lue

ofou

tsta

ndi

ng

equi

ty.W

eco

mpu

teR

etur

nVa

riab

ility

asth

ean

nua

lsta

nda

rdde

viat

ion

ofm

onth

lyst

ock

retu

rns.

Tota

lAss

ets

are

den

omin

ated

inU

S$m

illio

n.F

inan

cial

Lev

erag

eis

com

pute

das

the

rati

oof

tota

llia

bilit

ies

toto

tala

sset

s.Fo

reca

stB

ias

equa

lsth

eon

e-ye

ar-a

hea

dI/

B/E

/San

alys

tfor

ecas

terr

or(m

ean

fore

cast

min

usac

tual

)sc

aled

byla

gged

tota

lass

ets.

Infla

tion

isth

ean

nua

lmed

ian

ofth

eon

e-ye

ar-a

hea

dre

aliz

edm

onth

lype

rcen

tage

chan

ges

ina

coun

try’

sco

nsu

mer

pric

ein

dex

asre

port

edin

Dat

astr

eam

.Acc

oun

tin

gda

taan

dm

arke

tval

ues

are

mea

sure

das

ofth

efi

scal

-yea

ren

d,th

ede

pen

den

tvar

iabl

es,a

nal

ystf

ollo

win

g,fo

reca

stbi

as,r

etur

nva

riab

ility

,an

dsh

are

turn

over

asof

mon

th+1

0af

ter

the

end

ofth

efi

scal

year

.Exc

eptf

orva

riab

les

wit

hn

atur

allo

wer

orup

per

boun

ds,w

etr

unca

teal

lvar

iabl

esat

the

firs

tan

d99

thpe

rcen

tile

.

514 H. DASKE, L. HAIL, C. LEUZ, AND R. VERDIT

AB

LE

3R

egre

ssio

nA

naly

sis

ofL

iqui

dity

and

Cos

tofC

apita

lAro

und

Volu

ntar

yIA

SA

dopt

ions

Log

(Pri

ceIm

pact

)L

og(B

id-A

skSp

read

)C

osto

fCap

ital

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gVa

riab

les

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

IAS

−2.3

7−2

.45

0.51

3.45

∗−3

.36∗

5.02

∗∗∗

0.42

∗∗∗

0.20

0.33

∗∗

(−0.

83)

(−0.

81)

(0.1

8)(1

.79)

(−1.

84)

(2.6

0)(3

.29)

(1.3

0)(2

.26)

Rep

ortin

gVa

riab

leL

evel

−55.

96∗∗

∗−2

.53∗∗

∗−2

4.62

∗∗∗

−19.

68∗∗

∗−1

.38∗∗

∗−8

.90∗∗

∗−3

.26∗∗

∗−0

.13∗∗

∗−0

.43∗∗

∗

(−22

.17)

(−9.

56)

(−22

.62)

(−12

.54)

(−9.

60)

(−12

.62)

(−31

.94)

(−6.

80)

(−8.

33)

Con

trol

Vari

able

s:U

.S.G

AA

P5.

334.

585.

94−0

.71

0.16

2.89

−0.0

4−0

.40

−0.3

4(0

.90)

(0.7

8)(1

.00)

(−0.

20)

(0.0

5)(0

.82)

(−0.

12)

(−0.

99)

(−0.

87)

U.S

.Lis

ting

−25.

97∗∗

∗−2

6.68

∗∗∗

−25.

29∗∗

∗−2

.61

−6.8

0∗∗∗

−1.8

30.

18∗

−0.3

8∗∗∗

−0.3

5∗∗∗

(−10

.24)

(−10

.55)

(−9.

95)

(−1.

50)

(−4.

33)

(−1.

07)

(1.7

8)(−

3.41

)(−

3.19

)N

ewM

arke

ts21

.81∗∗

∗25

.54∗∗

∗30

.53∗∗

∗12

.19∗∗

∗14

.22∗∗

∗15

.35∗∗

∗0.

41∗

0.55

∗∗0.

59∗∗

(5.9

0)(6

.64)

(8.1

6)(5

.62)

(6.2

6)(7

.04)

(1.8

6)(2

.22)

(2.5

0)In

dex

Mem

ber

−56.

46∗∗

∗−6

2.08

∗∗∗

−51.

98∗∗

∗−1

8.49

∗∗∗

−20.

06∗∗

∗−1

6.68

∗∗∗

−0.7

3∗∗∗

−1.2

5∗∗∗

−1.1

1∗∗∗

(−30

.06)

(−30

.96)

(−27

.97)

(−14

.71)

(−16

.63)

(−13

.26)

(−9.

24)

(−14

.20)

(−13

.20)

Log

(Mar

ketV

alue

t−1)

−82.

89∗∗

∗−9

8.33

∗∗∗

−90.

71∗∗

∗−2

5.97

∗∗∗

−31.

15∗∗

∗−2

8.76

∗∗∗

––

–(−

93.9

3)(−

176.

20)

(−13

8.60

)(−

48.5

0)(−

94.5

0)(−

71.3

9)L

og(S

hare

Turn

over

t−1)

−66.

28∗∗

∗−6

6.86

∗∗∗

−63.

80∗∗

∗−2

1.53

∗∗∗

−20.

94∗∗

∗−2

0.84

∗∗∗

––

–(−

106.

96)

(−10

5.60

)(−

104.

09)

(−57

.98)

(−55

.41)

(−56

.48)

Log

(Ret

urn

Vari

abili

tyt−

1)

45.8

1∗∗∗

43.7

7∗∗∗

46.2

1∗∗∗

33.2

1∗∗∗

28.3

1∗∗∗

33.5

5∗∗∗

––

–(3

7.24

)(3

3.27

)(3

7.87

)(4

1.49

)(3

4.74

)(4

2.38

)L

og(T

otal

Ass

ets)

––

––

––

0.29

∗∗∗

−0.2

5∗∗∗

−0.1

2∗∗∗

(9.3

9)(−

9.07

)(−

4.13

)Fi

nanc

ialL

ever

age

––

––

––

3.92

∗∗∗

3.90

∗∗∗

3.43

∗∗∗

(21.

29)

(18.

88)

(17.

56)

Ret

urn

Vari

abili

ty–

––

––

–6.

25∗∗

∗6.

53∗∗

∗7.

11∗∗

∗

(10.

05)

(9.6

4)(1

1.21

)

(Con

tinue

d)


TA

BL

E3—

Con

tinue

d

Log

(Pri

ceIm

pact

)L

og(B

id-A

skSp

read

)C

osto

fCap

ital

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gR

epor

ting

Rep

ortin

gVa

riab

les

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

Ince

ntiv

esB

ehav

ior

Envi

ronm

ent

Fore

cast

Bia

s–

––

––

–19

.68∗∗

∗17

.98∗∗

∗18

.57∗∗

∗

(19.

38)

(17.

14)

(18.

62)

Infla

tion

––

––

––

28.0

6∗∗∗

24.8

0∗∗∗

26.5

5∗∗∗

(13.

19)

(10.

46)

(12.

03)

Cou

ntr

y-,Y

ear-

,an

dYe

sYe

sYe

sYe

sYe

sYe

sYe

sYe

sYe

sIn

dust

ry-F

ixed

Eff

ects

R2

81.4

%81

.9%

81.4

%73

.1%

75.9

%73

.3%

35.9

%30

.6%

30.4

%#

Obs

erva

tion

s68

,076

60,3

3969

,528

47,9

8542

,652

49,0

4325

,260

23,6

8925

,934

#U

niq

ueFi

rms

12,0

2410

,938

12,1

719,

722

8,82

59,

838

5,88

15,

439

6,05

1#

Cou

ntr

ies

3030

3024

2424

2929

29

Th

esa

mpl

eco

mpr

ises

firm

-yea

rob

serv

atio

ns

from

upto

30co

untr

ies

wit

hvo

lun

tary

IAS

adop

tion

sbe

twee

n19

90an

d20

05(s

eeta

ble

1).

Th

eta

ble

repo

rts

OL

Sco

effi

cien

tes

tim

ates

and

(in

pare

nth

eses

)t-

stat

isti

csba

sed

onro

bust

stan

dard

erro

rsth

atar

ecl

uste

red

byfi

rm.W

eus

ePr

iceI

mpa

ct,B

id-A

skSp

read

,an

dC

osto

fCap

ital

asth

ede

pen

den

tvar

iabl

es.

IAS

repr

esen

tsfi

rm-y

ears

wit

hfi

nan

cial

repo

rts

inac

cord

ance

wit

hIA

S/IF

RS,

base

don

our

augm

ente

dW

orld

scop

eac

coun

tin

gst

anda

rds

clas

sifi

cati

onde

scri

bed

inth

eap

pen

dix.

Inea

chre

gres

sion

,we

incl

ude

one

ofth

efo

llow

ing

thre

eR

epor

ting

Vari

able

s(c

ompu

ted

inye

art)

:(1)

the

stre

ngt

hof

firm

s’R

epor

ting

Ince

ntiv

es,(

2)fi

rms’

actu

alR

epor

ting

Beh

avio

r,or

(3)

the

exte

rnal

pres

sure

from

firm

s’R

epor

ting

Envi

ronm

ent.

For

vari

able

deta

ilsse

eta

bles

1an

d2.

We

use

the

nat

ural

log

ofth

era

wva

lues

and

lag

the

vari

able

sby

one

year

wh

ere

indi

cate

d.W

ein

clud

eco

untr

y-,y

ear-

,an

din

dust

ry-fi

xed

effe

cts

(bas

edon

the

clas

sifi

cati

onin

Cam

pbel

l[1

996]

)in

the

regr

essi

ons,

but

don

otre

port

the

coef

fici

ents

.For

expo

siti

onal

purp

oses

we

mul

tipl

yal

lcoe

ffici

ents

by10

0.∗∗

∗ ,∗∗

,an

d∗

indi

cate

stat

isti

cals

ign

ifica

nce

atth

e1%

,5%

,an

d10

%le

vels

(tw

o-ta

iled)

.


transparency, is associated with a strong reduction in price impact (in linewith, e.g., Foerster and Karolyi [1999] and Baruch, Karolyi, and Lemmon[2007]). Both the results for the reporting variables and U.S. cross-listingmake it unlikely that lack of power is responsible for the weak IAS results.There are also liquidity differentials for firms on new markets (consistentwith high growth, high risk firms); constituents of major stock indices (con-sistent with large, well established and less risky firms); and firms with largermarket valuations, higher share turnover, and higher return variability. Thecoefficient on voluntary U.S. GAAP is insignificant. In the model with firm-fixed effects (not tabulated), the IAS coefficient is significantly negative,providing a first indication that heterogeneity across firms matters.

Next, we present the bid-ask spread results. Two models suggest that firmsreporting under IAS have significantly larger spreads; the third model sug-gests the opposite. This may seem surprising, but it should be noted that intable 3 we include IAS firms for which we are unable to identify the switchyear (see table A4, panel B, in the appendix). Thus, the IAS coefficientcould also reflect cross-sectional differences between IAS and local GAAPfirms that are not fully controlled for in our specification. Consistent withthis explanation, the IAS coefficient becomes significantly negative in thefirm-fixed effects model (not tabulated).21 This finding underscores the im-portance of requiring a switch year in the subsequent analyses. The controlvariables exhibit associations similar to those in the price impact model.

The final three columns in table 3 report the cost of capital results. Sim-ilar to the spread regressions, two of the IAS coefficients are positive andsignificant, indicating that IAS firms on average have a higher cost of capitalthan local GAAP firms.22 Unlike for spreads, this result continues to holdin the firm-fixed effects specification (not tabulated). The three reportingvariables are always significantly negative, consistent with the idea that thesevariables capture differences in firms’ reporting incentives.23 Most of theremaining control variables are highly significant and have the expectedsigns.24

In sum, the three reporting variables exhibit significant associationsconsistent with the idea that these variables capture differences in firms’

21 Similarly, when we omit IAS firms without identifiable switch year, the IAS coefficientsacross the three bid-ask spread models are negative (but only significant in Model 2).

22 Daske [2006], examining a broad sample of German firms, also finds that voluntary IASadopters exhibit higher costs of equity capital than local GAAP firms.

23 This relation is also economically significant. For instance, a one-standard-deviation in-crease in the reporting variables implies a decrease in the cost of capital of 183 (ReportingIncentives), 33 (Reporting Behavior), and 43 basis points (Reporting Environment), respectively.More realistically, a 1% change in the reporting incentives variables implies a decrease in thecost of capital between 1 and 9 basis points.

24 The significantly positive coefficients on U.S. Listing and Total Assets in Model 1 are pri-marily due to the inclusion of contemporaneous market values when computing the ReportingIncentives variable (see also Hail and Leuz [2009]). When we replace market values with totalassets in this computation, we find a significantly negative coefficient for U.S. cross-listings.


reporting incentives. The average economic consequences associated withvoluntary IAS adoptions are mixed. One explanation for these mixed re-sults could be substantial heterogeneity across firms, in which case examin-ing the average effect can be misleading. We explore this possibility next.Before that, however, we note that our findings in table 3 are importantbecause they are inconsistent with the notion that voluntary IAS adoptionitself constitutes a commitment to increased disclosure. Such a commit-ment effect for the average or most firms would be surprising consideringthat firms have substantial flexibility in how they adopt IAS, and in light ofdescriptive evidence that there are major differences in the quality of IASfinancial statements (e.g., Cairns [1999, 2001], Barth, Landsman, and Lang[2008]).

4.2 HETEROGENEITY IN EFFECTS ACROSS SERIOUS AND LABEL ADOPTERS

We begin the analyses in this section with descriptive statistics for thethree reporting variables used to split the voluntary IAS firms into seriousand label adopters, and report them in table 4. In panel A, we tabulate thelevel of the reporting variables in the year immediately before the switchto IAS and the changes around IAS adoption that serve to create the bi-nary Serious Adopters indicators. The panel provides three insights. First, thepre-adoption level is substantially higher for label adopters than for seriousadopters. This indicates that label adopters have less room for improve-ment and are not necessarily firms with poor reporting practices to beginwith. Second, for serious adopters, the mean (median) change in all threereporting variables is positive, suggesting an increase in their incentives fortransparency. Yet, the average change for label adopters is negative. Third,the average changes are quite large, suggesting a substantial shift in re-porting incentives around IAS adoption. For instance, the mean changein Reporting Incentives equals –0.264 for label adopters and +0.359 for seri-ous adopters, relative to the pre-IAS level of 0.227. Table 4 also reports thedistribution of serious and label adopters by country (panel B) and year(panel C). The two types of firms occur in almost any country and no clus-tering is apparent. The number of serious adopters exceeds the number oflabel adopters in the early years, but the pattern reverses in the later years(except for 2004).

Next, in table 5, we provide median values and the number of observa-tions for the dependent variables across label and serious adopters, andlocal GAAP firms. We only include IAS firms with identifiable switch years(to distinguish between serious and label adopters) and “pure” local GAAPfirms (to control for time-period effects). As panels A and B show, seriousadopters have higher stock market liquidity than label adopters, which inturn have higher liquidity than local GAAP firms. For instance, in the Re-porting Incentives partition, the price impact for serious adopters is 0.046compared to 0.111 (0.185) for label adopters (local GAAP firms). Ex-cept in one case, the differences are all statistically significant. In panelC, we compare the cost of capital across groups. In the Reporting Incentives


TA

BL

E4

Des

crip

tive

Stat

istic

sfo

rSe

riou

san

dL

abel

Volu

ntar

yIA

SA

dopt

ions

Pan

elA

:Des

crip

tive

Stat

isti

csfo

rth

eR

epor

ting

Vari

able

sU

sed

toP

arti

tion

the

Volu

ntar

yIA

SFi

rms

into

Seri

ous

and

Lab

elA

dopt

ers

Vari

able

sN

Mea

nSt

d.D

ev.

P1M

edia

nP9

9Sp

litby

Cha

nges

inR

epor

ting

Ince

ntiv

es:

Lab

elA

dopt

ers

Pre-

adop

tion

Lev

el27

00.

330

0.50

1−1

.018

0.35

61.

314

�ar

oun

dA

dopt

ion

270

−0.2

640.

331

−1.5

82−0

.175

0.03

7Se

riou

sA

dopt

ers

Pre-

adop

tion

Lev

el25

70.

118

0.57

3−1

.198

0.09

41.

256

�ar

oun

dA

dopt

ion

257

0.35

90.

293

0.04

70.

277

1.41

0To

tal

Pre-

adop

tion

Lev

el52

70.

227

0.54

7−1

.163

0.22

01.

276

�ar

oun

dA

dopt

ion

527

0.04

00.

441

−1.3

330.

039

1.28

5

Split

byC

hang

esin

Rep

ortin

gB

ehav

ior:

Lab

elA

dopt

ers

Pre-

adop

tion

Lev

el24

1−0

.786

0.67

7−3

.903

−0.6

61− 0

.042

�ar

oun

dA

dopt

ion

241

−1.6

383.

703

−19.

852

−0.4

59−0

.004

Seri

ous

Ado

pter

sPr

e-ad

opti

onL

evel

229

−2.2

403.

139

−15.

395

−0.9

72−0

.197

�ar

oun

dA

dopt

ion

229

1.52

32.

922

0.02

00.

302

14.4

21To

tal

Pre-

adop

tion

Lev

el47

0−1

.494

2.35

6−1

4.65

4−0

.763

−0.0

97�

arou

nd

Ado

ptio

n47

0−0

.098

3.69

7−1

6.62

9−0

.003

13.5

87

Split

byC

hang

esin

Rep

ortin

gEn

viro

nmen

t:L

abel

Ado

pter

sPr

e-ad

opti

onL

evel

249

1.63

70.

920

0.23

11.

609

3 .21

9�

arou

nd

Ado

ptio

n24

9−0

.648

0.52

4−2

.193

−0.4

75−0

.024

Seri

ous

Ado

pter

sPr

e-ad

opti

onL

evel

338

0.59

40.

909

0.00

00.

000

3.11

9�

arou

nd

Ado

ptio

n33

80.

429

0.63

50.

000

0.08

62.

556

Tota

lPr

e-ad

opti

onL

evel

587

1.03

71.

049

0.00

00.

693

3.17

8�

arou

nd

Ado

ptio

n58

7−0

.028

0.79

5−2

.042

0.00

02.

533

(Con

tinue

d)


TA

BL

E4—

Con

tinue

d

Pan

elB

:Dis

trib

utio

nof

Seri

ous

and

Lab

elVo

lunt

ary

IAS

Ado

pter

sby

Cou

ntry

Split

byC

han

ges

inSp

litby

Ch

ange

sin

Split

byC

han

ges

inR

epor

tin

gIn

cen

tive

sR

epor

tin

gB

ehav

ior

Rep

orti

ng

En

viro

nm

ent

(Num

ber

ofU

niq

ueFi

rms)

Cou

ntr

yL

abel

Seri

ous

Tota

lL

abel

Seri

ous

Tota

lL

abel

Seri

ous

Tota

lA

ustr

alia

22

40

33

22

4A

ustr

ia22

1638

1714

3118

2543

Bel

gium

57

124

913

85

13B

erm

uda

02

20

11

02

2C

anad

a0

00

10

10

11

Ch

ina

2120

413

47

440

44C

zech

Rep

ublic

10

11

01

10

1D

enm

ark

97

1611

516

99

18Fi

nla

nd

23

52

46

53

8Fr

ance

719

267

2128

1118

29G

erm

any

8060

140

8366

149

6797

164

Gre

ece

41

52

24

33

6H

ong

Kon

g2

13

13

42

35

Hun

gary

22

43

25

33

6Is

rael

01

11

01

01

1It

aly

1519

3419

1029

1420

34L

uxem

bour

g2

13

30

30

33

Th

eN

eth

erla

nds

04

42

13

05

5N

orw

ay1

01

10

11

01

Paki

stan

44

82

57

35

8Pe

ru12

214

81

96

814

Pola

nd

20

22

13

31

4Po

rtug

al1

34

12

31

34

Rus

sian

Fede

rati

on1

34

03

31

45

Sout

hA

fric

a7

815

78

156

1016

Swed

en0

22

22

41

23

Swit

zerl

and

3643

7929

3867

3348

81Tu

rkey

3227

5929

2453

4717

64To

tal

270

257

527

241

229

470

249

338

587

(Con

tinue

d)


TA

BL

E4—

Con

tinue

d

Pan

elC

:Dis

trib

utio

nof

Seri

ous

and

Lab

elVo

lunt

ary

IAS

Ado

pter

sby

Year

Split

byC

han

ges

inSp

litby

Ch

ange

sin

Split

byC

han

ges

inR

epor

tin

gIn

cen

tive

sR

epor

tin

gB

ehav

ior

Rep

orti

ng

En

viro

nm

ent

(Num

ber

ofU

niq

ueFi

rms)

Cou

ntr

yL

abel

Seri

ous

Tota

lL

abel

Seri

ous

Tota

lL

abel

Seri

ous

Tota

l19

901

78

35

81

67

1991

32

52

24

05

519

925

1318

313

164

1822

1993

411

153

912

114

1519

945

914

67

135

712

1995

104

142

911

210

1219

966

915

26

87

815

1997

1313

269

716

1020

3019

9820

1131

1910

2916

2137

1999

4123

6438

2159

3443

7720

0045

2065

3921

6038

4179

2001

3314

4726

1642

2924

5320

0241

2667

3330

6327

4471

2003

3242

7431

3364

4636

8220

0411

5364

2540

6529

4170

Tota

l27

025

752

724

122

947

024

933

858

7%

inPr

e-19

98Pe

riod

17.4

%26

.5%

12.4

%25

.3%

12.0

%26

.0%

%in

Post

-199

7Pe

riod

82.6

%73

.5%

87.6

%74

.7%

88.0

%74

.0%

Th

esa

mpl

eco

mpr

ises

firm

-yea

rob

serv

atio

ns

from

upto

30co

untr

ies

wit

hvo

lun

tary

IAS

adop

tion

sbe

twee

n19

90an

d20

05(s

eeta

ble

1).T

he

tabl

ere

port

sde

scri

ptiv

est

atis

tics

for

the

thre

ere

port

ing

vari

able

sus

edto

part

itio

nth

evo

lun

tary

IAS

adop

tin

gfi

rms

into

seri

ous

and

labe

lad

opte

rs(p

anel

A),

and

the

num

ber

ofun

ique

(ser

ious

and

labe

l)IA

Sad

opte

rsby

coun

try

(pan

elB

)an

dye

ar(p

anel

C).

Inpa

nel

A,w

ere

port

desc

ript

ive

stat

isti

csfo

rth

efo

llow

ing

thre

eva

riab

les:

(1)

Th

est

ren

gth

offi

rms’

Rep

ortin

gIn

cent

ives

mea

sure

das

the

rolli

ng

thre

e-ye

arav

erag

eof

the

raw

scor

esfr

oma

sin

gle

fact

orex

trac

ted

from

vari

ous

firm

attr

ibut

es.(

2)Fi

rms’

actu

alR

epor

ting

Beh

avio

rm

easu

red

asth

ero

llin

gth

ree-

year

aver

age

ofth

eab

solu

teva

lue

ofac

crua

lssc

aled

byth

eab

solu

teva

lue

ofca

shfl

ows

from

oper

atio

ns.

(3)

Th

eex

tern

alpr

essu

refr

omfi

rms’

Rep

ortin

gEn

viro

nmen

tm

easu

red

asth

ero

llin

gth

ree-

year

aver

age

ofth

en

atur

allo

gof

the

num

ber

ofan

alys

tsfo

llow

ing

the

firm

(plu

son

e).T

he

pre-

adop

tion

leve

lis

equa

lto

the

resp

ecti

vere

port

ing

vari

able

inth

eye

arbe

fore

the

swit

chto

IAS

(i.e

.,ye

art

−1

rela

tive

toye

art

ofIA

Sad

opti

on).

Th

ech

ange

(�)

arou

nd

adop

tion

isth

edi

ffer

ence

betw

een

year

t+

3af

ter

adop

tion

and

year

t−

1be

fore

adop

tion

.We

then

part

itio

nth

eIA

Sad

opti

ng

firm

sba

sed

onth

edi

stri

buti

onof

thes

ech

ange

sin

toSe

riou

sA

dopt

ers

(ch

ange

isgr

eate

rth

anor

equa

lto

the

med

ian

)an

dL

abel

Ado

pter

s.E

ach

volu

nta

ryIA

Sad

opte

ren

ters

this

com

puta

tion

only

once

.Not

eth

at,d

ueto

addi

tion

alda

tare

stri

ctio

ns,

the

num

ber

ofun

ique

volu

nta

ryIA

Sfi

rms

islo

wer

inth

ela

ter

regr

essi

onan

alys

es.


TA

BL

E5

Uni

vari

ate

Ana

lysi

sof

Liq

uidi

tyan

dC

osto

fCap

italA

roun

dSe

riou

san

dL

abel

Volu

ntar

yIA

SA

dopt

ions

Spl

it b

y C

hang

es in

R

epor

ting

Inc

enti

ves

Spl

it b

y C

hang

es in

R

epor

ting

Beh

avio

rS

plit

by

Cha

nges

in

Rep

orti

ng E

nvir

onm

ent

(Med

ian

Val

ues

and

Num

ber

of

Obs

erva

tion

s)

Lab

el

Ado

pter

s(1

)

Ser

ious

A

dopt

ers

(2)

Dif

fere

nce

(2) –

(1)

Lab

el

Ado

pter

s(1

)

Ser

ious

A

dopt

ers

(2)

Dif

fere

nce

(2)–

(1)

Lab

el

Ado

pter

s(1

)

Ser

ious

A

dopt

ers

(2)

Dif

fere

nce

(2) –

(1)

Pan

elA

:Pri

ceIm

pact

0.11

10.

046

– –

– –

– –

– –

– 0.

101*

**

0.06

5***

0.10

70.

057

0.05

0***

0.09

00.

085

–0.0

05IA

S(a

)1,

060

880

874

754

889

1,26

2

0.18

50.

191

0.19

1L

ocal

GA

AP

(b

)63

,812

57,1

1165

,259

Dif

fere

nce

(a)

(b)

0.07

4***

0.13

9***

0.08

4***

0.13

4***

0.10

6***

Pan

elB

:Bid

-Ask

Spre

ad

0.01

20.

008

0.00

4***

0.01

20.

010

0.00

2***

0.01

10.

010

0.00

1***

IAS

(a)

877

717

762

653

771

989

0.01

90.

019

0.01

9L

ocal

GA

AP

(b

)44

,461

39,8

5745

,517

Dif

fere

nce

(a)

(b)

0.00

7***

0.01

1***

0.00

7***

0.00

9***

0.00

8***

0.00

9***

––

–

––

––

––

–

Pan

elC

:Cos

tof

Cap

ital

12.0

%10

.7%

––1

.3%

***

11.4

%11

.0%

0.4%

10.8

%11

.6%

0.8%

**IA

S(a

)39

343

739

444

033

057

6

11.1

%11

.0%

11.1

%L

ocal

GA

AP

(b

)23

,423

22,0

8024

,078

Dif

fere

nce

(a)

(b)

0.9%

***

0.4%

**0.

4%0.

0%0.

3%0.

5%*

––

–

Th

esa

mpl

eco

mpr

ises

firm

-yea

rob

serv

atio

ns

from

upto

30co

untr

ies

wit

hvo

lun

tary

IAS

adop

tion

sbe

twee

n19

90an

d20

05(s

eeta

ble

1).T

he

tabl

ere

port

sm

edia

nva

lues

and

the

num

ber

ofob

serv

atio

ns

acro

ssfi

rms

repo

rtin

gun

der

loca

lGA

AP,

seri

ous

and

labe

lIA

Sad

opte

rs.W

eus

ePr

ice

Impa

ct,B

id-A

skSp

read

,an

dC

osto

fCap

ital

asca

pita

l-mar

ket

vari

able

sin

the

anal

yses

(see

tabl

e2)

.IA

San

dL

ocal

GA

AP

repr

esen

tfi

rm-y

ears

wit

hfi

nan

cial

repo

rts

inac

cord

ance

wit

hei

ther

IAS/

IFR

Sor

dom

esti

cac

coun

tin

gst

anda

rds,

base

don

our

augm

ente

dW

orld

scop

eac

coun

tin

gst

anda

rds

clas

sifi

cati

onde

scri

bed

inth

eap

pen

dix.

We

clas

sify

each

IAS

adop

tin

gfi

rmas

eith

erSe

riou

sA

dopt

eror

Lab

elA

dopt

erba

sed

onth

edi

stri

buti

onof

the

chan

ges

inth

eth

ree

repo

rtin

gva

riab

les

arou

nd

IAS

adop

tion

(see

tabl

e4

for

deta

ils).

We

limit

the

IAS

obse

rvat

ion

sto

the

init

ialfi

veye

ars

afte

rth

ead

opti

on.∗∗

∗ ,∗∗

,an

d∗

indi

cate

stat

isti

cals

ign

ifica

nce

ofm

edia

ndi

ffer

ence

sat

the

1%,5

%,a

nd

10%

leve

ls(t

wo-

taile

d),b

ased

ona

Wilc

oxon

ran

ksu

mte

st.


partition, the serious adopters have lower costs of capital than local GAAPfirms; label adopters have the highest costs of capital. The results in theother two partitions are inconclusive (or opposite).

We then present OLS regression estimates of equation (1) in table 6.25

For price impact, we find that the coefficients on the Serious Adopters indica-tor are always negative and significant regardless of the partitioning variablewe use. That is, serious adopters are more liquid than label adopters and, asindicated by the significant F -test, more liquid than local GAAP firms. Thisfinding suggests substantial and predictable heterogeneity in the price im-pact effects, and is consistent with the notion that serious adopters increasetheir commitment to transparency around IAS adoption. The IAS coeffi-cient is significantly positive in the Reporting Incentives and Reporting Environ-ment partitions, indicating that label adopters exhibit lower market liquiditythan local GAAP firms. The control variables behave similarly to those intable 3. Notably, the three reporting variables (in levels) are still negativelyassociated with price impact.26 The fact that we find higher liquidity for se-rious adopters after controlling for the pre-IAS level of the respective parti-tioning variable suggests that our results are not driven by prior differencesin reporting incentives. It is also inconsistent with the notion that voluntaryadoptions lead to a convergence of reporting practices.

The bid-ask spread results closely resemble those for price impact. TheSerious Adopters coefficient is always significantly negative and so is the jointeffect with the IAS variable. Thus, serious adopters have smaller bid-askspreads compared to label adopters and local GAAP firms. For the labeladopters, the effect is significantly positive in two of the three regressions.To gauge the economic magnitude, we compare the post-adoption meanwith the pre-adoption mean spread under local GAAP (i.e., 0.033 in table 2,panel A). The coefficients in table 6 indicate an average decrease in spreadsfor serious adopters between 6.9% (Reporting Environment) and 11.4%(Reporting Incentives). Label adopters see a modest increase of 6.4% (Re-porting Incentives) or smaller. The spread differences between serious andlabel adopters range from 17.8% (Reporting Incentives) to 7.6% (ReportingBehavior).27 These effects are economically significant, as we would expectif these firms experience large changes in their reporting incentives and,in turn, make changes to their commitment to transparency.

For the cost of capital, we find that serious adopters exhibit a sig-nificantly lower cost of capital than label adopters in two of the three

25 Compared to table 3, the number of IAS firms is substantially lower (1) because we omitfirms without switch year, and (2) due to the required data to compute changes in the report-ing variables (see also table 4, panel A).

26 We compute the reporting variables for the respective year, except for IAS firms, forwhich we use the pre-IAS mean. Using the pre-IAS level in the panel regression prevents thesevariables from picking up the changes in reporting incentives around IAS adoption, which weuse to construct the serious versus label partitions.

27 The economic magnitudes are in a similar range for the other two dependent variables(i.e., slightly lower for the cost of capital and larger for price impact).


TA

BL

E6

Reg

ress

ion

Ana

lysi

sof

Liq

uidi

tyan

dC

osto

fCap

italA

roun

dSe

riou

san

dL

abel

Volu

ntar

yIA

SA

dopt

ions

Log

(Pri

ceIm

pact

)L

og(B

id-A

skSp

read

)C

osto

fCap

ital

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

Vari

able

sIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIA

S16

.95∗∗

∗6.

2818

.42∗∗

∗6.

17∗∗

−0.6

95.

40∗∗

1.34

∗∗∗

0.87

∗∗∗

0.46

∗

(3.5

1)(1

.25)

(3.5

5)(2

.19)

(−0.

25)

(1.9

8)(5

.21)

(2.8

6)(1

.67)

Seri

ous

Ado

pter

s−4

8.38

∗∗∗

−19.

28∗∗

∗−3

7.77

∗∗∗

−18.

33∗∗

∗−7

.98∗∗

−12.

53∗∗

∗−1

.78∗∗

∗−0

.68∗∗

−0.1

2(−

7.04

)(−

2.85

)(−

5.78

)(−

4.48

)(−

2.07

)(−

3.27

)(−

5.59

)(−

1.99

)(−

0.34

)IA

S+

Seri

ous=

0[p

-val

ue]

[0.0

0][0

.01]

[0.0

0][0

.00]

[0.0

0][0

.02]

[0.0

4][0

.33]

[0.1

4]C

ontr

olVa

riab

les:

Rep

ortin

gVa

riab

leL

evel

−56.

48∗∗

∗−2

.57∗∗

∗−2

4.73

∗∗∗

−18.

77∗∗

∗−1

.48∗∗

∗−8

.81∗∗

∗−3

.15∗∗

∗−0

.13∗∗

∗−0

.42∗∗

∗

(−22

.47)

(−9.

49)

(−22

.47)

(−11

.86)

(−10

.04)

(−12

.28)

(−29

.48)

(−6.

45)

(−8.

10)

U.S

.GA

AP

5.81

5.13

6.93

−4.3

9−0

.43

−0.0

4−0

.02

−0.3

6−0

.25

(0.9

6)(0

.86)

(1.1

5)(−

1.18

)(−

0.12

)(−

0.01

)(−

0.05

)(−

0.88

)(−

0.64

)U

.S.L

istin

g−2

4.96

∗∗∗

−26.

20∗∗

∗−2

4.54

∗∗∗

−3.5

1∗∗−6

.61∗∗

∗−2

.87∗

0.17

−0.3

7∗∗∗

−0.3

4∗∗∗

(−9.

58)

(−10

.16)

(−9.

40)

(−2.

03)

(−4.

15)

(−1.

68)

(1.6

3)(−

3.23

)(−

3.05

)N

ewM

arke

ts19

.43∗∗

∗23

.19∗∗

∗28

.57∗∗

∗13

.79∗∗

∗13

.37∗∗

∗16

.15∗∗

∗0.

48∗

0.51

∗0.

53∗∗

(4.8

5)(5

.74)

(7.3

5)(5

.69)

(5.4

1)(6

.95)

(1.8

7)(1

.90)

(2.0

3)In

dex

Mem

ber

−56.

76∗∗

∗−6

2.54

∗∗∗

−52.

27∗∗

∗−1

8.73

∗∗∗

−20.

02∗∗

∗−1

6.95

∗∗∗

−0.7

6∗∗∗

−1.2

7∗∗∗

−1.1

2∗∗∗

(−29

.79)

(−30

.69)

(−27

.81)

(−14

.85)

(−16

.28)

(−13

.27)

(−9.

38)

(−14

.22)

(−13

.19)

Log

(Mar

ketV

alue

t−1)

−83.

01∗∗

∗−9

8.36

∗∗∗

−90.

81∗∗

∗−2

6.16

∗∗∗

−31.

18∗∗

∗−2

8.70

∗∗∗

––

–(−

94.2

9)(−

172.

69)

(−13

7.04

)(−

49.0

0)(−

92.6

2)(−

70.1

9)L

og(S

hare

Turn

over

t−1)

−65.

84∗∗

∗−6

6.55

∗∗∗

−63.

47∗∗

∗−2

1.17

∗∗∗

−20.

76∗∗

∗−2

0.45

∗∗∗

––

–(−

103.

67)

(−10

2.98

)(−

101.

41)

(−57

.47)

(−54

.37)

(−54

.09)

Log

(Ret

urn

Vari

abili

tyt−

1)

45.6

1∗∗∗

43.7

0∗∗∗

46.0

5∗∗∗

33.0

6∗∗∗

28.3

8∗∗∗

33.3

4∗∗∗

––

–(3

6.35

)(3

2.33

)(3

7.01

)(4

0.22

)(3

4.59

)(4

0.71

)L

og(T

otal

Ass

ets)

––

––

––

0.27

∗∗∗

−0.2

6∗∗∗

−0.1

3∗∗∗

(8.4

5)(−

8.99

)(−

4.30

)Fi

nanc

ialL

ever

age

––

––

––

3.95

∗∗∗

3.88

∗∗∗

3.44

∗∗∗

(20.

88)

(18.

46)

(17.

28)

(Con

tinue

d)


TA

BL

E6—

Con

tinue

d

Log

(Pri

ceIm

pact

)L

og(B

id-A

skSp

read

)C

osto

fCap

ital

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

Vari

able

sIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tR

etur

nVa

riab

ility

––

––

––

6.28

∗∗∗

6.52

∗∗∗

7.00

∗∗∗

(9.7

2)(9

.38)

(10.

71)

Fore

cast

Bia

s–

––

––

–19

.23∗∗

∗17

.58∗∗

∗18

.31∗∗

∗

(18.

33)

(16.

32)

(17.

92)

Infla

tion

––

––

––

26.5

8∗∗∗

23.5

1∗∗∗

25.8

4∗∗∗

(11.

93)

(9.7

1)(1

1.37

)C

oun

try-

,Yea

r-,a

nd

Indu

stry

-Fix

edE

ffec

tsYe

sYe

sYe

sYe

sYe

sYe

sYe

sYe

sYe

s

R2

81.6

%81

.9%

81.5

%73

.5%

76.1

%73

.6%

35.6

%30

.6%

30.5

%#

Obs

erva

tion

s65

,752

58,7

3967

,390

46,0

5541

,272

47,2

7724

,253

22,9

1424

,984

#Se

riou

s/L

abel

Obs

erva

tion

s88

0/1,

060

754/

874

1,26

2/88

971

7/87

765

3/76

298

9/77

143

7/39

344

0/39

457

6/33

0#

Un

ique

Firm

s11

,544

10,5

7811

,753

9,28

78,

500

9,46

05,

574

5,20

55,

764

#C

oun

trie

s30

3030

2424

2429

2929

Th

esa

mpl

eco

mpr

ises

firm

-yea

rob

serv

atio

ns

from

upto

30co

untr

ies

wit

hvo

lun

tary

IAS

adop

tion

sbe

twee

n19

90an

d20

05(s

eeta

ble

1).

Th

eta

ble

repo

rts

OL

Sco

effi

cien

tes

tim

ates

and

(in

pare

nth

eses

)t-

stat

isti

csba

sed

onro

bust

stan

dard

erro

rsth

atar

ecl

uste

red

byfi

rm.I

tal

sore

port

sp-

valu

es[i

nbr

acke

ts]

from

anF

-test

indi

cati

ng

join

tst

atis

tica

lsi

gnifi

can

ceof

the

coef

fici

ents

onIA

San

dSe

riou

sA

dopt

ers.

We

use

Pric

eIm

pact

,Bid

-Ask

Spre

ad,a

nd

Cos

tofC

apita

las

the

depe

nde

nt

vari

able

s.IA

Sre

pres

ents

firm

-yea

rsw

ith

fin

anci

alre

port

sin

acco

rdan

cew

ith

IAS/

IFR

S,ba

sed

onou

rau

gmen

ted

Wor

ldsc

ope

acco

unti

ng

stan

dard

scl

assi

fica

tion

desc

ribe

din

the

appe

ndi

x.In

each

regr

essi

on,w

ein

clud

eon

eof

the

follo

win

gth

ree

Rep

ortin

gVa

riab

les

(com

pute

din

year

t,ex

cept

for

IAS

adop

tin

gfi

rms,

for

wh

ich

we

use

the

mea

nof

the

resp

ecti

vem

etri

cin

the

year

sbe

fore

IAS

adop

tion

):(1

)th

est

ren

gth

offi

rms’

Rep

ortin

gIn

cent

ives

,(2)

firm

s’ac

tual

Rep

ortin

gB

ehav

ior,

and

(3)

the

exte

rnal

pres

sure

from

firm

s’R

epor

ting

Envi

ronm

ent.

We

furt

her

clas

sify

each

IAS

adop

tin

gfi

rmas

eith

erse

riou

s(i

.e.,

Seri

ous

Ado

pter

sva

riab

lese

teq

ualt

o“1

”)or

labe

lado

pter

(i.e

.,Se

riou

sA

dopt

ers

vari

able

set

equa

lto

“0”)

base

don

the

dist

ribu

tion

ofth

ech

ange

s(�

)in

the

thre

ere

port

ing

vari

able

sar

oun

dIA

Sad

opti

on(s

eeta

ble

4fo

rde

tails

).W

elim

itth

eIA

Sob

serv

atio

ns

toth

ein

itia

lfive

year

saf

ter

the

adop

tion

.For

vari

able

deta

ils,s

eeta

bles

1an

d2.

We

use

the

nat

ural

log

ofth

era

wva

lues

and

lag

the

vari

able

sby

one

year

wh

ere

indi

cate

d.W

ein

clud

eco

untr

y-,y

ear-

,an

din

dust

ry-fi

xed

effe

cts

(bas

edon

the

clas

sifi

cati

onin

Cam

pbel

l[19

96])

inth

ere

gres

sion

s,bu

tdo

not

repo

rtth

eco

effi

cien

ts.F

orex

posi

tion

alpu

rpos

esw

em

ulti

ply

allc

oeffi

cien

tsby

100.

∗∗∗ ,

∗∗,a

nd

∗in

dica

test

atis

tica

lsig

nifi

can

ceat

the

1%,5

%,a

nd

10%

leve

ls(t

wo-

taile

d).


models. However, compared to local GAAP firms, the net effect is signif-icantly negative only in the Reporting Incentives partition. In the remain-ing cases, the combined coefficient is not distinguishable from zero. Thus,there is at best weak evidence suggesting a decline in the cost of capitalfor serious adopters. Furthermore, the IAS main effect is significantly posi-tive throughout, suggesting that label adopters have a higher cost of capitalthan local GAAP firms.

As discussed in section 2, there are several explanations for the positiveIAS coefficients in the liquidity and cost of capital regressions. First, it isimportant to recall that our analysis so far cautions us against interpret-ing the positive coefficients as attributable to label adoption per se. Theylikely reflect the negative changes in firms’ reporting incentives aroundIAS adoption, and hence a deterioration of transparency (see panel A oftable 4). In that sense, seeing adverse economic consequences around IASadoption is not surprising. That said, there are other possible explanations.For instance, label adoptions might not be that costly to begin with andfirms could obtain other (small) benefits (that we do not analyze), whichin turn make label adoptions worthwhile.28 Alternatively, managers couldperceive a general trend to IAS/IFRS and feel pressured to follow, eventhough they do not want to adjust their reporting practices by much.29 Wefurther note that the positive IAS coefficients are less robust than the neg-ative (incremental) effect of serious adoption (see also section 4.4).

In sum, the cross-sectional analyses confirm our main hypothesis thatthere is substantial heterogeneity in economic outcomes around IAS adop-tions, and that these differences predictably line up with the degree towhich firms exhibit substantial changes in their reporting incentives. Inaddition, the findings suggest that (at least some) investors can successfullydistinguish between different adoption types and do not reward all firmsirrespective of how they adopt IAS.

4.3 HETEROGENEITY IN THE STANDARDS AND MANDATORY IFRS ADOPTION

In this section, we explore the alternative explanation that heteroge-neous effects around voluntary IAS adoptions reflect heterogeneity in the

28 It could be that IAS adoption is a public relations tool for these firms (e.g., in communi-cations with foreigners) and that it has an impact on foreign holdings, but less so on liquidityas one group of investors simply replaces another (e.g., Covrig, DeFond, and Hung [2007]).See also Tan, Wang, and Welker [2011] for related results.

29 Consistent with such a “bandwagon” effect, label adoptions become relatively more fre-quent as the trend toward IAS accelerates (see table 4, panel C). We also note that there area large number of serious adopters in 2004. These firms are likely stronger firms that volun-tarily adopt ahead of the IFRS mandate to signal their type (see also Daske et al. [2008]).In addition, when we reestimate our main regressions separately for the subperiods pre andpost the completion of the IASC’s Core Standards Project in 1997 (not tabulated), we findthat the main effect of the IAS indicator variable is generally more positive in the post-1997period. This is what one would expect if these firms are compelled by the overall trend towardIAS/IFRS.


standards that are being adopted. The set of IAS has changed considerablyover the years. Similarly, firms do not always fully adopt IAS. Such hetero-geneity across time and firms could lead to differential effects across vol-untary IAS adoptions, and hence could be a potential explanation for ourfindings.

We conduct three sets of tests to gauge the importance of this expla-nation: first, we reestimate our analyses using a very narrow definition ofwhat constitutes voluntary IAS, thereby reducing the heterogeneity in thestandards that voluntary IAS adopters follow. Second, we apply the seriousversus label distinction to firms around the mandatory switch to IFRS. Inprinciple, our reporting incentives argument should also apply to manda-tory IFRS adopters, in particular considering that the mandate forces firmsto adopt IFRS, and hence there are likely some firms that would have pre-ferred to continue reporting under local GAAP. At the same time, hetero-geneity in the standards is less of an issue for mandatory IFRS adoption.Thus, if our results for voluntary adoptions were driven primarily by hetero-geneity in the standards, we should not obtain similar or at least consider-ably weaker results around mandatory adoptions. Third, we study voluntaryU.S. GAAP adoptions, for which there has been a complete set of account-ing standards for the entire sample period. We present the results fromthe first two tests in this section. The voluntary U.S. GAAP results follow insection 4.4.

Table 7, panel A, reports liquidity and cost of capital results across seri-ous and label adopters using a very strict (and narrow) definition of whatconstitutes IAS reporting. That is, starting with the augmented Worldscopeaccounting standards classification, we impose two additional criteria: first,we limit the sample period to the years 1998 to 2005. In 1998, the IASCfinished a major revision of its core standards and from this point onwardIAS were considered a complete set of standards. The revised standardsalso eliminated many optional accounting treatments. Second, we requirean explicit auditor attestation of compliance with IAS or IFRS. To do so, wemanually check whether the auditor’s report contains a statement like “inour opinion, the consolidated financial statements . . . are in accordancewith the IAS as issued by the IASB.”30

As shown in the table, using a stricter IAS definition nearly cuts the sam-ple of IAS firms in half, and we lose several countries for which there areno longer any IAS adopters. Nonetheless, the results remain very similar tothose before. Even for this narrow sample, there is no evidence that IASadoption on average improves liquidity or lowers the cost of capital. Forboth liquidity variables, the Serious Adopters coefficient (net effect) is neg-ative and generally significant, indicating that serious adopters have lower

30 To avoid a reduction in sample size due to gaps in our panel of annual reports, we re-quire such a statement within the first three years after IAS adoption. However, the results arequalitatively the same if we limit the analyses to IAS firms with an auditor’s attestation in theactual year of the switch.


TA

BL

E7

Sens

itivi

tyA

naly

ses:

Het

erog

enei

tyin

the

Acc

ount

ing

Stan

dard

san

dM

anda

tory

IFR

SA

dopt

ion

Pan

elA

:Cap

ital

Mar

ketE

ffec

tsA

roun

dSe

riou

san

dL

abel

Volu

ntar

yIA

SA

dopt

ions

for

Stri

ctIA

SD

efini

tion

Aft

er19

98(R

equi

ring

Aud

itor

Att

esta

tion

)L

og(P

rice

Impa

ct)

Log

(Bid

-Ask

Spre

ad)

Cos

tofC

apita

l

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

Vari

able

sIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIA

S19

.05∗∗

∗8.

1714

.40∗∗

12.9

0∗∗∗

4.77

10.2

7∗∗∗

1.45

∗∗∗

1.23

∗∗∗

0.71

∗∗

(3.1

4)(1

.31)

(2.2

5)(3

.53)

(1.4

2)(3

.03)

(4.6

2)(3

.24)

(2.2

1)Se

riou

sA

dopt

ers

−48.

04∗∗

∗−1

8.69

∗∗−3

6.18

∗∗∗

−23.

79∗∗

∗−1

0.88

∗∗−1

6.73

∗∗∗

−1.9

3∗∗∗

−1.5

4∗∗∗

−0.5

1(−

5.23

)(−

2.17

)(−

4.45

)(−

4.59

)(−

2.32

)(−

3.45

)(−

4.50

)(−

3.40

)(−

1.09

)IA

S+

Seri

ous=

0[0

.00]

[0.1

2][0

.00]

[0.0

1][0

.10]

[0.1

0][0

.14]

[0.2

9][0

.57]

[p-v

alue

]C

ontr

olVa

riab

les,

Cou

ntr

y-,

Year

-,an

dIn

dust

ry-F

ixed

Eff

ects

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

R2

81.1

%81

.5%

81.1

%76

.2%

78.8

%76

.4%

42.4

%37

.8%

37.3

%#

Obs

erva

tion

s32

,102

29,7

9432

,823

24,2

0422

,474

24,7

588,

117

7,91

08,

453

#Se

riou

s/L

abel

Obs

erva

tion

s44

4/57

040

2/51

755

8/57

638

3/53

238

7/48

449

3/53

615

8/21

419

2/19

322

7/19

6#

Un

ique

Firm

s7,

654

7,12

67,

793

6,07

45,

657

6,19

02,

605

2,45

62,

696

#C

oun

trie

s25

2525

2020

2023

2323

(Con

tinue

d)


TA

BL

E7—

Con

tinue

d

Pan

elB

:Cap

ital

Mar

ketE

ffec

tsA

roun

dSe

riou

san

dL

abel

Man

dato

ryIF

RS

Ado

ptio

nsL

og(P

rice

Impa

ct)

Log

(Bid

-Ask

Spre

ad)

Cos

tofC

apita

l

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

Mod

el1:

Mod

el2:

Mod

el3:

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

�R

epor

ting

Vari

able

sIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIn

cent

ives

Beh

avio

rEn

viro

nmen

tIF

RS

39.9

5∗∗∗

20.3

9∗∗∗

37.1

6∗∗∗

13.9

0∗∗∗

8.78

∗∗∗

9.41

∗∗∗

0.25

0.05

0.20

(10.

64)

(5.2

0)(8

.38)

(7.0

9)(4

.59)

(4.0

5)(1

.57)

(0.3

0)(1

.05)

Seri

ous

Ado

pter

s−6

7.53

∗∗∗

−19.

44∗∗

∗−3

8.43

∗∗∗

−25.

83∗∗

∗−1

2.65

∗∗∗

−10.

63∗∗

∗−1

.11∗∗

∗−0

.19

−0.4

3∗∗

(−14

.84)

(−4.

09)

(−7.

94)

(−10

.48)

(−5.

06)

(−4.

05)

(−5.

81)

(−0.

95)

(−2.

17)

IFR

S+

Seri

ous=

0[p

-val

ue]

[0.0

0][0

.80]

[0.6

9][0

.00]

[0.0

5][0

.45]

[0.0

0][0

.41]

[0.1

0]

Con

trol

Vari

able

s,C

oun

try-

,Yea

r-,a

nd

Indu

stry

-Fix

edE

ffec

ts

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

R2

85.6

%85

.7%

85.9

%77

.2%

78.1

%77

.6%

36.5

%33

.9%

33.4

%#

Obs

erva

tion

s22

,421

21,3

7523

,964

21,7

0920

,726

23,2

806,

897

6,96

07,

260

#Se

riou

s/L

abel

Obs

erva

tion

s1,

551/

1,51

61,

540/

1,55

12,

443/

983

1,49

3/1,

465

1,49

5/1,

499

2,38

4/93

451

0/50

255

5/46

764

7/42

3

#U

niq

ueFi

rms

6,82

56,

491

7,25

26,

590

6,25

87,

012

2,80

62,

801

2,94

0#

Cou

ntr

ies

2626

2623

2323

2323

23T

he

sam

ple

com

pris

esfi

rm-y

ear

obse

rvat

ion

sfro

mup

to25

coun

trie

swit

hvo

lun

tary

IAS

adop

tion

sbet

wee

n19

98an

d20

05in

pan

elA

,an

dfr

omup

to26

coun

trie

swit

hm

anda

tory

IFR

Sad

opti

onby

the

end

of20

05(c

over

ing

the

year

s20

02to

2005

)in

pan

elB

.Th

eta

ble

repo

rts

only

the

OL

Sco

effi

cien

ts(a

nd

inpa

ren

thes

est-

stat

isti

cs,c

lust

ered

byfi

rm)

ofth

eIA

San

dSe

riou

sA

dopt

ers

(IFR

San

dSe

riou

sA

dopt

ers)

vari

able

s,bu

tth

efu

llse

tof

con

trol

san

dfi

xed

effe

cts

isin

clud

ed.I

tal

sore

port

sp-

valu

es[i

nbr

acke

ts]

from

anF

-test

indi

cati

ng

join

tsta

tist

ical

sign

ifica

nce

ofth

etw

ova

riab

les.

We

use

Pric

eIm

pact

,Bid

-Ask

Spre

ad,a

nd

Cos

tofC

apita

las

the

depe

nde

ntv

aria

bles

.In

pan

elA

,IA

Sre

pres

ents

firm

-yea

rsw

ith

fin

anci

alre

port

sin

acco

rdan

cew

ith

IAS/

IFR

S,ba

sed

onou

rau

gmen

ted

Wor

ldsc

ope

acco

unti

ng

stan

dard

scl

assi

fica

tion

desc

ribe

din

the

appe

ndi

x.In

addi

tion

,we

requ

ire

the

follo

win

gtw

ocr

iter

iato

ensu

repr

oper

IAS/

IFR

Sre

port

ing

byth

etr

eatm

ent

firm

s:(1

)w

elim

itth

esa

mpl

eto

firm

-yea

rsfr

om19

98on

war

d,i.e

.,th

eye

arth

eIA

SCco

mpl

eted

am

ajor

revi

sion

ofal

lits

core

stan

dard

s,an

d(2

)w

eon

lyin

clud

eIA

Sad

opti

ng

firm

sfo

rw

hic

hco

mpl

ian

cew

ith

IAS/

IFR

Sre

port

ing

inth

eth

ree

year

sfo

llow

ing

the

swit

chis

atte

sted

byth

eau

dito

rin

the

audi

top

inio

nfo

otn

ote

(bas

edon

our

man

uali

nsp

ecti

onof

the

firm

’sfi

nan

cial

repo

rts)

.In

pan

elB

,IFR

Sre

pres

ents

firm

-yea

rsen

din

gon

oraf

ter

the

loca

ldat

efo

rm

anda

ted

IFR

Sre

port

ing

(see

Das

keet

al.

[200

8]),

and

appl

ies

only

tofi

rms

that

nev

erre

port

edun

der

IAS/

IFR

Sbe

fore

(i.e

.,w

eex

clud

evo

lun

tary

IAS

adop

ter

firm

s).

We

use

the

sam

eth

ree

Rep

ortin

gVa

riab

les

asco

ntr

ols

and

topa

rtit

ion

the

man

dato

ryIF

RS

adop

tin

gfi

rms

into

seri

ous

and

labe

lado

pter

sas

inth

evo

lun

tary

IAS

anal

yses

.For

deta

ilson

the

regr

essi

onsp

ecifi

cati

onan

dth

ese

riou

sve

rsus

labe

lcla

ssifi

cati

on,s

eeta

ble

6.Fo

rva

riab

lede

tails

,see

tabl

es1

and

2.Fo

rex

posi

tion

alpu

rpos

esw

em

ulti

ply

allc

oeffi

cien

tsby

100.

∗∗∗ ,

∗∗,a

nd

∗in

dica

test

atis

tica

lsig

nifi

can

ceat

the

1%,5

%,a

nd

10%

leve

ls(t

wo-

taile

d).


price impact and bid-ask spreads than label adopters (local GAAP firms).The cost of capital results are similar to table 6, except that serious adoptersare not different from local GAAP firms in a statistical sense.

Next, we apply our partitioning technique to mandatory IFRS adoption.We limit the sample to the subset of countries from table 1 that switchedto IFRS reporting for consolidated financial statements by the end of 2005,and only cover the years 2002 to 2005. We define a binary IFRS indicator forfirms that are required to adopt IFRS for the first time, and use it in placeof the IAS variable. Consequently, we identify the market effects relativeto the firms’ own pre-adoption years and the firms in the same countriesthat have not yet or did not adopt IFRS (i.e., we eliminate voluntary IASfirms from the analyses). We split the mandatory IFRS adopters into seriousand label firms using the same partitioning technique as before. For eachreporting variable we compute the change around the mandated switch(thereby utilizing data up to 2008), and assign the Serious Adopters indicatorto the firms with above-median changes. The rest of the model specificationis the same as in equation (1).

Panel B of table 7 reports the results for liquidity and cost of capitalaround mandatory IFRS adoption. Throughout, the Serious Adopters coef-ficient is negative and, with one exception, significant. Hence, firms withrelatively large increases in reporting incentives exhibit higher liquidity andlower cost of capital than firms that adopt IFRS only in name. In five out ofthe nine specifications, market liquidity (cost of capital) is also significantlyhigher (lower) for serious adopters than for local GAAP firms. Moreover,price impact and bid-ask spreads are higher for label adopters than for lo-cal GAAP firms. Such a positive coefficient on IFRS is not surprising if wecorrectly identify firms that were forced to adopt IFRS, but would have pre-ferred local GAAP (see also Christensen, Lee, and Walker [2007]). Thiscould also explain why the adverse effects for label adoptions appear to belarger around the mandate than around voluntary adoption.

Overall, the evidence around mandatory adoptions is similar to the ev-idence around voluntary adoptions. This similarity makes it unlikely thatheterogeneity in the standards across firms or time is the primary driverof our results. More importantly, the evidence in this section suggests thatreporting incentives also matter for market outcomes around changes inmandatory reporting.

4.4 SENSITIVITY ANALYSES

In this section, we test the sensitivity of our results to various researchdesign choices and report results in table 8. For brevity, we tabulate onlythe coefficients and t-statistics of the two IAS variables but, unless statedotherwise, estimate the regressions using the same models as in table 6.First, in panel A, we extend the analysis to three additional economic out-comes, namely Total Trading Costs (equal to the round trip transaction costsdrawn from Lesmond, Ogden, and Trzcinka [1999]), Zero Returns (equal tothe yearly proportion of trading days without stock price movements), and


T A B L E 8Additional Sensitivity Analyses for the Capital Market Effects Around Serious and Label Voluntary

IAS Adoptions

Split by Split by Split by� Reporting � Reporting � Reporting

Variables Incentives Behavior Environment

Panel A: Alternative Dependent Variables

(1) Log(Total Trading Costs)IAS −9.11∗∗∗ −4.82 −6.49∗

(−2.93) (−1.62) (−1.77)Serious Adopters −7.80∗ −7.57∗ −10.60∗∗

(−1.87) (−1.72) (−2.48)IAS + Serious = 0 [p-value] [0.00] [0.00] [0.00]

(2) Zero ReturnsIAS −2.49∗∗∗ −2.16∗∗∗ −1.57∗∗

(−4.05) (−3.26) (−2.15)Serious Adopters −0.22 −1.19 −2.03∗∗


(3) Tobin’s qIAS −36.02∗∗∗ −5.94 −17.24∗∗∗

(−12.12) (−1.34) (−3.99)Serious Adopters 57.17∗∗∗ 10.12 20.73∗∗∗

(11.64) (1.55) (3.34)IAS + Serious = 0 [p-value] [0.00] [0.43] [0.47]

Panel B: Alternative Accounting Standards Classifications (Price Impact as DependentVariable)

(4) Voluntary IAS and U.S. GAAP Adoption CombinedIAS/U.S. GAAP 16.31∗∗∗ 5.45 20.51∗∗∗

(3.53) (1.10) (4.11)Serious Adopters −48.19∗∗∗ −18.76∗∗∗ −39.72∗∗∗

(−7.42) (−2.95) (−6.45)IAS/U.S. GAAP + Serious = 0

[p-value][0.00] [0.01] [0.00]

(5) Voluntary U.S. GAAP AdoptionU.S. GAAP 11.03 −6.07 18.02

(0.83) (−0.39) (1.31)Serious Adopters −56.48∗∗∗ −16.90 −42.14∗∗

(−2.97) (−0.94) (−2.56)U.S. GAAP + Serious = 0 [p-value] [0.00] [0.05] [0.04]

Panel C: Alternative Sample Compositions (Price Impact as Dependent Variable)

(6) Exclude Firms from the United Kingdom and CanadaIAS 13.54∗∗∗ 1.17 14.82∗∗∗

(2.79) (0.23) (2.79)Serious Adopters −46.64∗∗∗ −16.19∗∗ −38.22∗∗∗


(7) IAS Firms OnlyIAS 4.63 −4.91 2.70

(0.82) (−0.87) (0.46)Serious Adopters −44.69∗∗∗ −11.70∗ −28.94∗∗∗


(Continued)


T A B L E 8—Continued

Split by Split by Split by� Reporting � Reporting � Reporting

Variables Incentives Behavior Environment

Panel D: Alternative Model Specifications (Price Impact as Dependent Variable)

(8) Firm-Fixed Effects Instead of Country- and Industry-Fixed EffectsIAS 20.19∗∗ −1.98 −2.79

(2.41) (−0.24) (−0.29)Serious Adopters −82.36∗∗∗ −24.08∗ −20.75


The sample comprises firm-year observations from up to 30 countries with voluntary IAS adoptionsbetween 1990 and 2005 (see table 1). The table reports only the OLS coefficients (and in parentheses t-statistics, clustered by firm) of the IAS and Serious Adopters (U.S. GAAP and Serious Adopters) variables, butthe full set of controls and fixed effects is included (see table 6). It also reports p-values [in brackets] froman F -test indicating joint statistical significance of the two variables. We show results for the serious andlabel adopters using each of the three Reporting Variables to partition the IAS (U.S. GAAP) observations. Inpanel A, we examine the following alternative dependent variables: Total Trading Costs is a yearly estimateof total round-trip transaction costs (i.e., bid-ask spreads, commissions, and implicit costs such as short-sale constraints or taxes) inferred from the time-series of daily security and aggregate market returns (seeLesmond, Ogden, and Trzcinka [1999]). Zero Returns is the proportion of trading days with zero daily stockreturns out of all potential trading days in a given year. Tobin’s q equals (total assets – book value of equity+ market value of equity)/total assets. The total trading costs and zero returns specifications are the sameas for price impact. For Tobin’s q we use the log of total assets, financial leverage, asset growth (computedas the one-year percentage change in total assets), and industry q (equal to the yearly median q in a givenindustry) as continuous control variables. In the remaining panels, we use Price Impact as the dependentvariable. In panel B, we report results for two alternative accounting standards classifications. First, wecreate a combined IAS and U.S. GAAP variable (based on our augmented Worldscope accounting standardsclassification), and split the firms that switch from local GAAP to either IAS or U.S. GAAP into serious andlabel adopters. Second, we focus on the voluntary U.S. GAAP adopters and partition them into serious andlabel adopters (while controlling for IAS reporting in the regression). In panel C, we report results foralternative sample compositions. First, we exclude firms from the United Kingdom and Canada from thesample. Second, we limit the analysis to firms that at some point during the sample period reported underIAS. In panel D, we replace the country- and industry-fixed effects with firm-fixed effects. For expositionalpurposes we multiply all coefficients by 100. ∗∗∗, ∗∗, and ∗ indicate statistical significance at the 1%, 5%, and10% levels (two-tailed).

firm value defined as Tobin’s q. For a description of the variables and regres-sion models, see the notes to table 8. The results using total trading costsare very similar to the liquidity results presented above. Serious adoptershave lower trading costs than label adopters. The results are weaker us-ing zero return days, but point in the same direction. In both cases, thenet effect for serious adopters is significantly negative. The effect for labeladopters is negative and significant in five out of six cases, indicating thatthe “adverse” effects of label adoptions shown in table 6 depend on theoutcome variable and are not very robust. The Tobin’s q results are con-sistent with the cost of capital effects. In particular, serious adopters havehigher firm values than label adopters and, in one case, than local GAAPfirms.31

31 Based on the mean pre-adoption Tobin’s q under local GAAP, serious adopters increasetheir q, on average, between 2.3% (Reporting Environment) and 13.7% (Reporting Incentives).Label adopters see a decrease between 3.9% (Reporting Behavior) and 23.4% (Reporting Incen-tives). These effects are economically significant and quite large (i.e., larger than the average


Second, in panel B, we evaluate whether the results are confined toIAS/IFRS.32 We extend the analysis by including voluntary U.S. GAAP adop-tions by non-U.S. firms (that are not cross-listed in the United States).Combining IAS and U.S. GAAP or, alternatively, considering only voluntaryU.S. GAAP adoptions yields essentially the same insights as before. Mar-ket liquidity is highest for serious adopters, while the differences betweenlocal GAAP firms and label adopters are usually not significant. These re-sults highlight that our findings for the reporting partitions are not spe-cific to voluntary IAS adoptions. Next, in tests not tabulated, we rerun ouranalyses using (1) the original Worldscope classification, (2) the originalGlobal Vantage classification, and (3) the subsample of firms with manu-ally coded accounting standards. We find that serious adopters have signifi-cantly higher levels of liquidity compared to label adopters and local GAAPfirms, regardless of the accounting standards classification. This result mayseem surprising in light of the many misclassifications across data sourcesdocumented in the appendix. However, because our partitioning is inde-pendent of the accounting standards classification, it is also possible thatmany misclassifications end up in the label adopter category, for which wedo not predict significant market reactions.

Third, in panel C, we gauge the effects of sample composition. We rees-timate the models excluding observations from the two countries with thelargest contribution to the overall sample (i.e., the United Kingdom andCanada). Both countries did not allow IAS reporting for statutory purposes,and hence they have a small number of voluntary IAS adopters. Next, welimit the sample to only firms that actually switch to IAS reporting (reduc-ing the sample to about 4,000 firm-years). The results indicate that ouranalyses remain largely unaffected by the differing sample definitions.33

One concern about the research design is that several of our control vari-ables as well as the inputs for the Reporting Incentives and Reporting Behaviorvariables use accounting numbers and that these numbers might changemechanically as a result of the switch to a different set of accounting stan-dards. With regard to this concern, we first note that the models that com-bine liquidity proxies with the partition based on changes in analyst follow-ing do not rely on any accounting data. Table 6 shows that the results forthese specifications are similar to the other models. Thus, we do not thinkthat mechanical effects from the change in accounting standards induce

effects of mandatory IFRS adoption in Daske et al. [2008], but smaller than the effects ofU.S. cross-listing in Doidge, Karolyi, and Stulz [2004]). But again, these effects should notbe attributed to IAS/IFRS alone as they likely also reflect more fundamental changes (whichhappen to be correlated with IAS adoption).

32 From here on, we report only the results for price impact as the dependent variable inthe table, but the inferences are similar using bid-ask spreads and cost of capital.

33 In further tests (not tabulated), we confirm that our results hold when we eliminateall sample restrictions (increasing the sample to 50 countries and a maximum of 135,538firm-years), and when we use a random benchmark sample consisting of up to 150 non-IASadopting firms per country and year.


our findings. Moreover, in panel D of table 8, when we reestimate the cross-sectional analyses using firm-fixed effects and eliminate time-invariant un-observed firm attributes as potentially confounding factors, the results con-tinue to hold, and the inferences are essentially the same as in our mainanalysis.

Finally, in untabulated analyses, we explore the interplay between firm-level incentives and country-level institutional factors, which could act assubstitutes or complements. For instance, in countries with strong capital-market incentives, one would expect firms to already report in a transpar-ent fashion (under local GAAP) and hence, when they switch to IAS, theydo not need to improve their reporting. In contrast, firms in countries withweak institutions could seek to commit to more transparency (e.g., becausethey face new growth opportunities), and hence they adopt a series of mea-sures to improve transparency, one of which is IAS. As shown in panel Bof table 4, we do not observe any clustering of serious and label adoptionsin certain countries. To assess this relation more formally, we run threeadditional tests. First, we construct two-by-two contingency tables for seri-ous and label adopters with two commonly used country-level institutionalfactors, each split by the median (i.e., the Djankov, La Porta, and Lopez-de-Silanes [2008] anti-self dealing index, and the La Porta et al. [1997]rule of law index). We generally find no statistical relation. Second, wecompute the Pearson correlation coefficients between the proportion ofserious adopters per country and the two country-level factors. The correla-tions are not significant. Third, we partition the sample countries into twogroups and rerun our main regression analyses. We find that the hetero-geneity in the capital-market effects around IAS adoption is present withincountries with weak and strong institutions and not limited to a particulargroup of countries. These results are consistent with firm-level incentiveshaving separate effects around IAS adoption that go beyond the effects ofcountry-level institutions.

5. Conclusion

This paper examines the economic consequences associated with volun-tary and mandatory IAS/IFRS adoptions around the world. We focus onfirm-level heterogeneity in the consequences, recognizing that firms candiffer in their motivations and ways to adopt IAS/IFRS. Some firms mayadopt new standards merely in name without making material changesto their reporting policies, while others adopt them as part of a broaderstrategy to strengthen their commitment to transparency. The possibility ofsuch differences implies significant heterogeneity in the economic conse-quences around IAS/IFRS adoptions due to selection effects.

To show the existence of such effects, we create three binary parti-tions that attempt to capture major changes in firms’ reporting incentivesaround voluntary IAS adoptions (based on various firm attributes, firms’actual reporting behavior, and the external pressure from the reporting


environment). We then examine the economic consequences around IASadoptions for “label” and “serious” firms in a large global panel from 1990to 2005.

We find little evidence that voluntary IAS adoptions are, on average, associ-ated with an increase in market liquidity or a decline in the cost of capital.If anything, the effects go in the opposite direction. However, using ourthree partitions, we document substantial differences in the capital marketeffects around IAS adoptions and find that concurrent changes in firms’reporting incentives play a significant role in explaining them. Specifically,serious adopters exhibit economically and statistically lower price impactof trades, bid-ask spreads, and costs of capital relative to label adopters andlocal GAAP firms. Given that we use changes in the incentives variables topartition IAS adoptions and include the corresponding pre-IAS level as acontrol in the model, our results are unlikely to reflect prior differences inreporting incentives, reporting quality, or the reporting environment.

Our evidence is inconsistent with the notion that IAS reporting per seconstitutes a commitment to increased transparency. Consequently, onehas to exercise caution in attributing capital market effects around vol-untary IAS adoptions solely to the change in accounting standards. Theobserved effects are more likely to reflect changes in firms’ reporting in-centives, including those that give rise to the decision to adopt IAS in thefirst place. One implication is that one could find, on average, positive, neg-ative, or no capital-market effects around IAS adoptions depending on thecomposition of the sample, unless such differences are accounted for. Assuch, our results help explain the mixed findings in prior literature.

Our study also contributes to the literature on the role of incentives forreporting outcomes and their capital-market effects. It highlights the roleof firm-level incentives (in addition to or separate from country-level in-centives) around voluntary IAS adoptions. The findings extend to volun-tary U.S. GAAP adoptions and, more importantly, to mandatory IFRS adop-tions. Furthermore, our study shows that markets can differentiate betweenserious and label adoptions. This is important as there is considerable con-cern that the global movement toward a single set of accounting standardsmasks the heterogeneity in actual reporting practices and makes it harderfor investors to evaluate firms’ reporting quality. Our results do not supportthis concern.

Finally, we caution the reader that our results should not be interpretedas indicating that “serious” adopters experience the documented effectsbecause they strictly comply with IAS/IFRS. Our partitions merely indicatethat these firms experience substantial changes in their reporting incen-tives around IAS/IFRS adoption, which likely influences their reportingbehavior more generally. For such firms, the market reaction likely reflectschanges in the entire reporting or commitment strategy, not just the switchto IAS/IFRS. Thus, a key message of our paper is that, in order to isolatethe marginal effects of IAS/IFRS reporting, one has to push harder on theidentification of such effects and, in particular, needs to separate them from


reporting incentive effects. For this very reason, our results should not beread as suggesting that standards do not matter. While we show that report-ing incentives play an important role for the sign and magnitude of themarket reactions around IAS/IFRS adoptions, our tests are not designed toanalyze the relative contribution of standards and incentives. We leave thisissue to future research.

APPENDIX

Classification of Voluntary IAS Adoptions Around the World

This appendix describes our coding of firm-year observations from firmsfollowing IAS/IFRS, U.S. GAAP, or local GAAP, and compares accountingstandards classifications across different data sources.34 It also documentsfirms’ reporting patterns across countries and over time, and how we iden-tify the year of voluntary IAS adoption.

We use three main data sources to construct our panel of firms’ account-ing standards: (1) Thomson Financial Worldscope (WS), (2) CompustatGlobal Vantage (GV), and (3) an extensive manual review of annual re-ports collected through Thomson Research. WS serves as a starting pointbecause its coverage is by far the most comprehensive. In addition, it is di-rectly linked to Datastream, thereby reducing the potential for mismatchesfrom combining accounting data with price data. Using the informationon accounting standards followed (field 07536), we classify firm-year obser-vations into the three categories IAS, U.S. GAAP, and local GAAP accord-ing to panel A of table A1. To assess the quality of commercially availabledatabases, we also classify the same set of observations applying the GV cod-ing scheme in panel B of table A1 that is based on the accounting standardsinformation in field “ASTD.”35

We next conduct an extensive manual data collection and classification.We start by identifying potential voluntary IAS or U.S. GAAP adopters, i.e.,firms with at least one firm-year coded as IAS or U.S. GAAP in either WSor GV before the end of 2005 (35,633 firm-years). We then gather thetime-series of annual reports through Thomson Research, and were ableto collect 22,213 documents in electronic image format. Based on a man-ual review of the accounting principles’ footnote and the auditors’ report,

34 We further illustrate our hand-coding, providing examples of each case usingexcerpts from firms’ annual reports, in an online appendix (available at http://research.chicagobooth.edu/arc/journal/onlineappendices.aspx). On this site, we also makethe database of voluntary IAS reporting created for this paper available for download.

35 One of the drawbacks of commercial databases is that they attempt to capture many dif-ferent reporting practices around the world, but often at the expense of consistency throughtime or across countries. Furthermore, the categories often do not provide clear distinctionsbetween local GAAP and IAS, as is needed for our study. For instance, category 02 “Interna-tional standards” in WS comprises not only IAS observations, but also firms following othernonlocal, non-U.S. standards (e.g., H.K. GAAP, U.K. GAAP). This problem is even more pro-nounced in GV as it has only three categories dedicated to international standards.


we create our own accounting standards classification. We deliberately usea broad classification when coding the binary IAS variable. Given our re-search question, we intend to capture a wide variety of adoption strategies,including dual reporting, reconciliations of local GAAP numbers to IAS, orfirms merely creating the appearance of IAS reporting. That is, our classi-fication relies primarily on what firms claim they do. This leads to six re-porting categories ranging from the exclusive use of IAS for consolidatedfinancial statements to reporting under local GAAP together with a rec-onciliation of net income and/or shareholders’ equity to IAS (table A1,panel C). Importantly, for our manual coding, we require that the annualreport contains an explicit reference to IAS/IFRS (U.S. GAAP) for the finan-cial statements in the footnotes. We classify firms as local when they adoptonly individual IAS or U.S. GAAP standards (e.g., for leasing or segmentreporting). Finally, we complete firms’ time-series by filling in cases withmissing individual annual reports, utilizing all data sources at hand (i.e.,WS, GV, annual reports). This procedure results in a total hand-coded sam-ple of 27,589 firm-year observations.

The main purpose of this massive hand-collection is to assess the suitabil-ity of commercially available databases for our research question. To gaugethe effect of potential misclassifications, we tabulate firm-years across differ-ent classifications and report the number of observations and percentagesin table A2. For ease of comparison, we use the base sample from our mainanalyses (n = 69,528). In panel A, we compare the classifications across WSand GV. For our comprehensive WS coding (i.e., when using all categorieslisted in table A1), we find that 2.4% of all firm-years are coded as IAS in WSbut as local GAAP in GV (compared to 1.2% the other way round). Thus,the two data sources provide contradictory information on about every thirdIAS firm-year observations (i.e., [7 + 844]/2,852 ≈ 30% when starting withGV; [26 + 1,658]/3,685 ≈ 46% when starting with WS). When we limit theWS coding to categories 02 and 23 and the GV coding to “DI” (labeled“Stricter Coding for IAS” in panel A), the contradiction rates remain sub-stantial (i.e., [7 + 921]/2,728 ≈ 34% for GV; [1 + 737]/2,538 ≈ 29% forWS). Panel B reports results from comparing the hand-coded classificationto WS and GV. Neither commercial database clearly dominates, but bothexhibit substantial classification differences relative to our manual reviewof annual reports, as indicated by large numbers of observations off the di-agonals. Our hand-coding disagrees in about 25% of the cases with WS orGV, computed as the fraction of off-diagonal observations relative to the ob-servations for which we have annual report data and WS (GV) information.Panel C reports Pearson correlations between the classifications.

As a further validity check, we compare the three coding schemes on acountry-by-country basis. In table A3 we tabulate the numbers and propor-tions of IAS and U.S. GAAP firms for all countries with a minimum of 20observations and total assets available in a given year. The table highlightsthe larger coverage in WS, which identifies more than twice as many IASfirm-years than GV (13,001 vs. 6,227). The hand-coded sample consists of


8,399 IAS firm-years. It further reveals that the proportion of IAS firms atthe individual country-level varies substantially. For instance, the percent-age of IAS adopters in Italy is as high as 78.7% according to WS but only0.2% based on GV. The hand-coding, to which we added observations thatare classified as local under both WS and GV, indicates a proportion of25.1%.

As a result of the above validity checks, we conduct our main analysesusing an “augmented” WS accounting standards classification, for whichwe use our hand-coded data together with the “Accounting Standard” fieldin GV to triangulate and correct the initial WS coding. This leads to 2,202cases for which our hand-coded data overrides conflicting WS information.Unless indicated otherwise, the results presented in the study rely on theaugmented WS coding to identify the IAS firm-year observations.36

The final step involves identifying the switch year. As table A4 illustrates,voluntary IAS adoption and reporting over time has taken many differentforms and patterns. In its simplest form, a firm starts out reporting in ac-cordance to local GAAP and then switches to IAS. In more complicatedcases, firms switch back and forth between local GAAP and IAS multipletimes, and also might switch to reporting in accordance with U.S. GAAP,for instance, as a result of a U.S. cross-listing. We assign IAS switch yearsonly if we can manually confirm that out of two consecutive firm-years thefirst is classified as local GAAP and the second as IAS.37 Consequently, weidentify 845 firms with proper IAS switch years (panel A), and exclude 570IAS reporting firms from the serious versus label partitions because we areunable to identify a switch year according to our definition (panel B).38

36 For sensitivity purposes, we also rerun the analyses with the pure Worldscope, GlobalVantage, or hand-coded classifications. In addition, we conduct tests using only IAS firms forwhich we can manually confirm IAS/IFRS reporting via explicit audit firm attestation in theannual report. See sections 4.3 and 4.4 for details.

37 We manually ensure that the splicing of different data sources does not introduce arti-ficial IAS adoption years. Moreover, we do not consider a switch from U.S. GAAP to IAS ascreating a valid switch year.

38 We check that our results are not confounded by the many variations in firms’ IAS adop-tion patterns (not tabulated). Even in the most restrictive case, when we limit the group of IASadopters to the 696 firms with only one switch from local GAAP to IAS, the inferences of ouranalyses remain unaffected.


T A B L E A 1Accounting Standards Classifications Based on Different Data Sources

Panel A: Coding Based on Worldscope (WS) “Accounting Standards Followed” (Field 07536)Coding for

WS Code WS Description AnalysesWe code firm-year observations as IAS if one of the following cases applies: IAS02 International standards06 International standards and some EU guidelines08 Local standards with EU and IASC guidelines12 International standards – inconsistency problems16 International standards and some EU guidelines – inconsistency

problems18 Local standards with some IASC guidelines19 Local standards with OECD and IASC guidelines23 IFRS

We code firm-year observations as U.S. GAAP if one of the following cases applies: U.S. GAAP03 U.S. standards (GAAP)13 U.S. standards – inconsistency problems20 U.S. GAAP reclassified from local standards

We code firm-year observations as local if one of the following cases applies: Local01 Local standards05 EU standards07 Specific standards set by the group09 Not disclosed10 Local standards with some EU guidelines11 Local standards – inconsistency problems14 Commonwealth standards – inconsistency problems15 EEC standards – inconsistency problems17 Local standards with some OECD guidelines21 Local standards with a certain reclassification for foreign companies22 OtherPanel B: Coding Based on Global Vantage (GV) “Accounting Standard” (Field ASTD)

Coding forGV Code GV Description AnalysesWe code firm-year observations as IAS if one of the following cases applies: IASDA Domestic standards generally in accordance with IASC and OECD

guidelinesDI Domestic standards generally in accordance with IASC guidelinesDT Domestic standards in accordance with principles generally accepted in

the United States and generally in accordance with IASC and OECDguidelines

We code firm-year observations as U.S. GAAP if one of the following cases applies: U.S. GAAPDU Domestic standards in accordance with principles generally accepted in

the United StatesMU Modified United States’ standards (Japanese companies’ financial

statements translated into English)US United States’ standards

We code firm-year observations as local if one of the following cases applies: LocalDD Domestic standards for parents and domestic subsidiaries. Native country

or United States’ standards for overseas subsidiariesDO Domestic standards generally in accordance with OECD guidelinesDR Accounts reclassified to show allowance for doubtful accounts and/or

accumulated depreciation as a reduction of assets rather than liabilitiesDS Domestic standardsMI Accounts reclassified by SPCS to combine separate life insurance and

nonlife insurance accountsLJ Combination DR and MI

(Continued)


T A B L E A 1—Continued

Panel C: Hand-Coded Classification Based on Firms’ Annual ReportsCoding for

Description AnalysesWe code firm-year observations as IAS if one of the following cases applies: IAS(1) Notes to consolidated financial statements refer to IAS/IFRS only(2) Annual report has two separate sections with two full sets of consolidated

financial statements (balance sheet, income statement, statement ofcash flows), one set under local GAAP, and one set under IAS/IFRS(Parallel Reporting)

(3) Notes to consolidated financial statements refer to IAS/IFRS in the firstplace, but also refer to compliance with local GAAP

(4) Notes to consolidated financial statements refer to local GAAP in the firstplace, but also refer to compliance with IAS/IFRS

(5) Notes to consolidated financial statements refer to full compliance withlocal GAAP, but also to application of IAS/IFRS if local GAAP is silentabout a reporting issue (Dual Reporting)

(6) Notes to consolidated financial statements refer to full application of localGAAP, but there is also a reconciliation of net income and/orshareholders’ equity to IAS/IFRS in a separate section of the annualreport (Reconciliation)

We code firm-year observations as U.S. GAAP if one of the following casesapplies:

U.S. GAAP

(1) Notes to consolidated financial statements refer to U.S. GAAP only(2) Annual report has two separate sections with two full sets of consolidated

financial statements (balance sheet, income statement, statement ofcash flows), one set under local GAAP, and one set under U.S. GAAP(Parallel Reporting)

(3) Notes to consolidated financial statements refer to U.S. GAAP in the firstplace, but also refer to compliance with local GAAP

(4) Notes to consolidated financial statements refer to local GAAP in the firstplace, but also refer to compliance with U.S. GAAP

(5) Notes to consolidated financial statements refer to full compliance withlocal GAAP, but also to application of U.S. GAAP if local GAAP is silentabout a reporting issue (Dual Reporting)

(6) Notes to consolidated financial statements refer to full application of localGAAP, but there is also a reconciliation of net income and/orshareholders’ equity to U.S. GAAP in a separate section of the annualreport (Reconciliation)

We code firm-year observations as local if one of the following cases applies: Local(1) Notes to consolidated financial statements refer to local GAAP only(2) Notes to consolidated financial statements refer to local GAAP only, but

selected individual IAS/IFRS or U.S. GAAP standards are applied onspecific reporting issues (e.g., Leasing IAS 17, Segment Reporting SFAS131)

The table describes the assignment of firm-year observations to the reporting categories IAS, U.S. GAAPor Local GAAP using three different accounting standards classifications. The first two are based on World-scope (panel A) and Global Vantage (panel B), as retrieved in May 2006. The third classification is hand-coded and based on firms’ annual reports collected through Thomson Research (panel C). In our mainanalyses, we use an augmented classification that combines information from all three classifications.


AB

LE

A2

Com

pari

son

ofA

ccou

ntin

gSt

anda

rds

Cla

ssifi

catio

nsby

Dat

aSo

urce

(for

Sam

ple

Use

din

Mai

nA

naly

ses)

Pan

elA

:Wor

ldsc

ope

Vers

usG

loba

lVan

tage

Cla

ssifi

cati

ons

Com

preh

ensi

veC

odin

gfo

rIA

SSt

rict

erC

odin

gfo

rIA

S(W

orld

scop

eC

lass

ifica

tion

)(W

orld

scop

eC

lass

ifica

tion

)

Glo

balV

anta

geU

.S.

U.S

.C

lass

ifica

tion

IAS

GA

AP

Loc

alTo

tal

IAS

GA

AP

Loc

alTo

tal

IAS

2,00

17

844

2,85

21,

800

792

12,

728

2.9%

0.0%

1.2%

4.1%

2.6%

0.0%

1.3%

3.9%

U.S

.GA

AP

2660

617

881

01

481

7856

00.

0%0.

7%0.

5%1.

2%0.

0%0.

7%0.

1%0.

8%L

ocal

1,65

821

845

,273

47,1

4973

733

246

,454

47,5

232.

4%0.

5%64

.9%

67.8

%1.

1%0.

5%66

.8%

68.4

%N

otco

vere

d66

225

217

,803

18,7

1744

725

218

,018

18,7

170.

9%0.

4%25

.6%

26.9

%0.

6%0.

4%25

.9%

26.9

%To

tal

4,34

71,

083

64,0

9869

,528

2,98

51,

072

65,4

7169

,528

6.3%

1.6%

92.2

%10

0.0%

4.3%

1.5%

94.2

%10

0.0%

Pan

elB

:Han

d-C

oded

Cla

ssifi

cati

onVe

rsus

Wor

ldsc

ope

and

Glo

balV

anta

geC

lass

ifica

tion

sW

orld

scop

eC

lass

ifica

tion

Glo

balV

anta

geC

lass

ifica

tion

Han

d-co

ded

U.S

.N

otU

.S.

Not

Cla

ssifi

cati

onIA

SG

AA

PL

ocal

Cov

ered

Tota

lIA

SG

AA

PL

ocal

Cov

ered

Tota

l

IAS

2,91

266

725

n.a

.3,

703

2,15

06

1,00

953

83,

703

4.2%

0.1%

1.0%

5.3%

3.1%

0.0%

1.5%

0.8%

5.3%

U.S

.GA

AP

4578

737

7n

.a.

1,20

920

537

406

246

1,20

90.

1%1.

1%0.

5%1.

7%0.

0%0.

8%0.

6%0.

4%1.

7%L

ocal

923

663,

288

n.a

.4,

277

399

692,

981

828

4,27

71.

3%0.

1%4.

7%6.

2%0.

6%0.

1%4.

3%1.

2%6.

2%N

oan

nual

repo

rtda

ta46

722

359

,649

n.a

.60

,339

283

198

42,7

5317

,105

60,3

39

0.7%

0.3%

85.8

%86

.8%

0.4%

0.3%

61.5

%24

.6%

86.8

%To

tal

4,34

71,

083

64,0

98n

.a.

69,5

282,

852

810

47,1

4918

,717

69,5

286.

3%1.

6%92

.2%

100.

0%4.

1%1.

2%67

.8%

26.9

%10

0.0%

(Con

tinue

d)


TA

BL

EA

2—C

ontin

ued

Pan

elC

:Pea

rson

’sC

orre

lati

onC

oeffi

cien

tsof

IAS

Cla

ssifi

cati

ons

Acr

oss

Dat

aSo

urce

sW

orld

scop

eG

loba

lVan

tage

Cla

ssifi

cati

onC

lass

ifica

tion

Han

d-co

ded

Cla

ssifi

cati

on0.

609

0.60

5p-

valu

es(0

.00)

(0.0

0)N

9,17

67,

570

Wor

ldsc

ope

Cla

ssifi

cati

on0.

593

p-va

lues

(0.0

0)N

50,7

75

Th

eta

ble

pres

ents

the

num

ber

ofob

serv

atio

ns

and

perc

enta

ges

acro

ssdi

ffer

enta

ccou

nti

ng

stan

dard

scl

assi

fica

tion

sfo

rou

rba

sesa

mpl

eco

mpr

isin

g69

,528

firm

-yea

rob

serv

atio

ns

from

30co

untr

ies

betw

een

1990

and

2005

(see

tabl

e1)

.In

pan

elA

,w

eco

mpa

reIA

S,U

.S.

GA

AP,

orlo

cal

GA

AP

firm

-yea

rsba

sed

onth

eac

coun

tin

gst

anda

rds

clas

sifi

cati

onin

Wor

ldsc

ope

and

Glo

balV

anta

ge.T

he

left

-han

dsi

deof

the

pan

elap

plie

sth

eco

din

gsc

hem

eas

outl

ined

inta

ble

A1.

On

the

righ

t-han

dsi

dew

eus

ea

stri

cter

codi

ng

for

IAS

obse

rvat

ion

sco

nsi

stin

gof

Wor

ldsc

ope

cate

gori

es02

(“In

tern

atio

nal

stan

dard

s”)

and

23(“

IFR

S”)

toge

ther

wit

hG

loba

lVa

nta

geca

tego

ryD

I(“

Dom

esti

cst

anda

rds

gen

eral

lyin

acco

rdan

cew

ith

IASC

guid

elin

es”)

.In

pan

elB

,we

com

pare

our

own

clas

sifi

cati

onba

sed

onan

exte

nsi

vem

anua

lre

view

offi

rms’

ann

ual

repo

rts

(see

tabl

eA

1,pa

nel

C)

wit

hth

eco

din

gba

sed

onW

orld

scop

ean

dG

loba

lVan

tage

.Pan

elC

repo

rts

the

resp

ecti

vePe

arso

nco

rrel

atio

ns

acro

ssth

eth

ree

data

sour

ces.


AB

LE

A3

Com

pari

son

ofA

ccou

ntin

gSt

anda

rds

Cla

ssifi

catio

nsby

Cou

ntry

Wor

ldsc

ope

Glo

balV

anta

geH

and-

Cod

edC

lass

ifica

tion

Cla

ssifi

cati

onC

lass

ifica

tion

plus

Loc

alG

AA

PFi

rms

IAS

U.S

.GA

AP

IAS

U.S

.GA

AP

IAS

U.S

.GA

AP

Cou

ntr

yn

n%

n%

nn

%n

%n

n%

n%

Arg

enti

na

899

30.

310

1.1

361

00.

06

1.7

337

30.

97

2.1

Aus

tral

ia12

,371

142

1.1

220.

24,

917

170.

38

0.2

4,04

413

43.

323

0.6

Aus

tria

2,02

733

216

.444

2.2

1,18

328

223

.838

3.2

1,24

534

327

.643

3.5

Bel

gium

3,02

218

76.

239

1.3

1,63

613

58.

348

2.9

1,34

117

913

.355

4.1

Ber

mud

a31

550

15.9

120

38.1

4,42

226

86.

186

1.9

220

5123

.210

547

.7B

razi

l3,

673

20.

16

0.2

1,69

80

0.0

20.

11,

374

10.

11

0.1

Can

ada

17,5

0287

0.5

418

2.4

9,64

113

0.1

240.

27,

669

400.

552

96.

9C

hile

2,06

31

0.0

20.

11,

106

00.

01

0.1

998

00.

00

0.0

Ch

ina

10,5

981,

144

10.8

108

1.0

14,0

6622

41.

64

0.0

8,51

51,

157

13.6

540.

6C

olom

bia

527

00.

00

0.0

265

00.

00

0.0

197

00.

00

0.0

Cze

chR

epub

lic47

090

19.1

00.

019

473

37.6

00.

020

668

33.0

00.

0D

enm

ark

3,85

414

13.

714

0.4

2,08

611

85.

70

0.0

1,85

315

58.

423

1.2

Egy

pt14

52

1.4

00.

00

00.

00

0.0

00

0.0

00.

0Fi

nla

nd

2,69

770

2.6

00.

01,

359

957.

03

0.2

1,29

879

6.1

80.

6Fr

ance

15,5

3396

16.

218

21.

28,

100

591

7.3

200

2.5

5,52

834

76.

319

73.

6G

erm

any

15,7

441,

685

10.7

714

4.5

8,21

21,

632

19.9

841

10.2

7,68

11,

723

22.4

679

8.8

Gre

ece

3,13

343

1.4

290.

974

238

5.1

162.

262

941

6.5

233.

7H

ong

Kon

g8,

855

200

2.3

590.

72,

374

130.

52

0.1

5,79

316

32.

833

0.6

Hun

gary

380

192

50.5

71.

818

213

272

.58

4.4

225

154

68.4

00.

0In

dia

4,54

78

0.2

170.

42,

644

40.

26

0.2

2,01

00

0.0

180.

9In

don

esia

3,26

81

0.0

00.

02,

541

00.

07

0.3

2,20

40

0.0

70.

3Ir

elan

d1,

498

402.

730

2.0

784

10.

16

0.8

660

00.

027

4.1

Isra

el1,

264

453.

636

128

.664

621

3.3

112

17.3

786

496.

241

152

.3It

aly

5,54

44,

364

78.7

80.

12,

944

60.

218

0.6

3,23

681

325

.134

1.1

Japa

n56

,817

133

0.2

944

1.7

45,1

845

0.0

348

0.8

36,6

670

0.0

637

1.7

Kor

ea(S

outh

)8,

566

100.

15

0.1

2,14

60

0.0

20.

11,

987

00.

05

0.3

Lux

embo

urg

484

9920

.550

10.3

305

9330

.539

12.8

191

107

56.0

2513

.1

(Con

tinue

d)


TA

BL

EA

3—C

ontin

ued

Wor

ldsc

ope

Glo

balV

anta

geH

and-

Cod

edC

lass

ifica

tion

Cla

ssifi

cati

onC

lass

ifica

tion

Plus

Loc

alG

AA

PFi

rms

IAS

U.S

.GA

AP

IAS

U.S

.GA

AP

IAS

U.S

.GA

AP

Cou

ntr

yn

n%

n%

nn

%n

%n

n%

n%

Mal

aysi

a9,

123

240.

30

0.0

8,16

734

0.4

00.

06,

718

20.

00

0.0

Mex

ico

1,99

37

0.4

80.

497

78

0.8

30.

386

210

1.2

151.

7M

oroc

co14

01

0.7

00.

00

00.

00

0.0

00

0.0

00.

0T

he

Net

her

lan

ds4,

295

185

4.3

232

5.4

2,53

395

3.8

162

6.4

2,15

113

16.

119

28.

9N

ewZ

eala

nd

1,46

331

2.1

00.

075

612

1.6

253.

362

96

1.0

284.

5N

orw

ay3,

120

351.

163

2.0

1,72

223

1.3

271.

61,

411

302.

167

4.7

Paki

stan

1,01

716

1.6

00.

045

412

2.6

00.

043

268

15.7

00.

0Pe

ru81

210

1.2

40.

50

00.

00

0.0

283

149

52.7

62.

1Ph

ilipp

ines

2,11

72

0.1

180.

91,

575

134

8.5

100.

678

97

0.9

212.

7Po

lan

d88

082

9.3

50.

636

259

16.3

00.

024

356

23.0

52.

1Po

rtug

al1,

368

332.

41

0.1

551

213.

80

0.0

351

3610

.30

0.0

Rus

sian

Fede

rati

on38

711

329

.281

20.9

150

6040

.051

34.0

220

9141

.457

25.9

Sin

gapo

re5,

465

147

2.7

350.

64,

410

170.

422

0.5

3,99

613

93.

539

1.0

Sout

hA

fric

a5,

936

751.

34

0.1

1,74

014

98.

65

0.3

1,45

216

211

.24

0.3

Spai

n3,

313

220.

77

0.2

1,90

68

0.4

50.

31,

556

60.

45

0.3

SriL

anka

256

00.

00

0.0

00

0.0

00.

00

00.

00

0.0

Swed

en5,

376

174

3.2

00.

03,

099

132

4.3

220.

72,

651

933.

514

0.5

Swit

zerl

and

4,86

31,

515

31.2

130

2.7

3,11

31,

294

41.6

105

3.4

3,20

71,

356

42.3

117

3.6

Taiw

an8,

873

10.

011

0.1

2,47

70

0.0

00.

02,

249

00.

015

0.7

Th

aila

nd

5,03

35

0.1

00.

04,

140

90.

21

0.0

3,53

04

0.1

00.

0Tu

rkey

1,92

125

513

.32

0.1

469

156

33.3

102.

194

527

629

.26

0.6

Un

ited

Kin

gdom

37,7

0023

30.

683

0.2

18,2

2856

0.3

500.

315

,669

170

1.1

530.

3Ve

nez

uela

454

30.

717

3.7

00

0.0

00.

00

00.

00

0.0

Tota

l29

1,70

113

,001

4.5

3,89

01.

317

8,45

96,

227

3.5

2,34

51.

314

6,23

88,

399

5.7

3,58

82.

5

Th

esa

mpl

eco

mpr

ises

all

firm

-yea

rob

serv

atio

ns

inW

orld

scop

ean

dG

loba

lVa

nta

gefr

omco

untr

ies

wit

hat

leas

t20

obse

rvat

ion

san

dto

tal

asse

tsav

aila

ble

ina

give

nye

ar.

Th

eh

and-

code

dcl

assi

fica

tion

isba

sed

on27

,589

obse

rvat

ion

sfr

omfi

rms

that

Wor

ldsc

ope

orG

loba

lVan

tage

iden

tify

aspo

ten

tial

IAS/

IFR

Sor

U.S

.GA

AP

user

san

dfo

rw

hic

hw

ew

ere

able

toco

llect

ann

ual

repo

rts

thro

ugh

Th

omso

nR

esea

rch

.We

then

add

all

obse

rvat

ion

sof

firm

sth

atar

eun

ifor

mly

clas

sifi

edas

repo

rtin

gun

der

loca

lG

AA

Pby

both

Wor

ldsc

ope

and

Glo

bal

Van

tage

toth

eh

and-

code

dsa

mpl

e.T

he

tabl

ere

port

sth

eto

tal

num

ber

offi

rm-y

ears

and

the

num

ber

and

perc

ent

ofIA

Sor

U.S

.GA

AP

obse

rvat

ion

s,us

ing

the

thre

eac

coun

tin

gst

anda

rds

clas

sifi

cati

ons

outl

ined

inta

ble

A1.


T A B L E A 4IAS Adoption Patterns and Identification of IAS Switch Years

Description of IAS Adoption Patterns over TimeUnique Firms

Firm-Years %

Panel A: Firms with IAS Switch Year

In the following (stylized) cases, we can identify a switch year to IAS/IFRS reporting, and hence classify firms as serious or label IAS adopters for our analyses.

Years: 1 2 3 4 5 6 7 8

(1) Firms switching from local GAAP to IASLocal Local Local Local IAS IAS IAS IAS

696 4,483 6.5

(2) Firms switching from local GAAP to IAS, and then back to local GAAPLocal Local IAS IAS IAS IAS Local Local

87 422 0.6

(3) Firms switching from local GAAP to IAS, back to local GAAP, and then for a second time to IASLocal Local IAS IAS Local Local IAS IAS

19 71 0.1

(4) Firms switching twice from local GAAP to IAS, and back to local GAAPLocal IAS IAS Local Local IAS IAS Local

1 3 0.0

(5) Firms switching from IAS to local GAAP, and then back to IASIAS IAS Local Local Local IAS IAS IAS

34 206 0.3

(6) Firms switching from IAS to local GAAP, back to IAS, and then for a second time to local GAAP

IAS IAS Local Local IAS IAS Local Local

6 37 0.1

(7) Firms switching from local GAAP to IAS, and then to U.S. GAAPLocal Local Local IAS IAS IAS U.S. U.S.

2 6 0.0

Panel B: Firms Without IAS Switch YearIn the following (stylized) cases, we are unable to identify a switch year to IAS/IFRSreporting, and hence exclude the firms from the serious versus label IAS partitions.

Years: 1 2 3 4 5 6 7 8

(1) Firms reporting under IAS during the entire time periodIAS IAS IAS IAS IAS IAS IAS IAS

544 2,691 3.9

(2) Firms switching from IAS to local GAAPIAS IAS IAS IAS Local Local Local Local

26 144 0.2

(3) Firms switching from local GAAP to U.S. GAAP, and then to IASLocal Local U.S. U.S. U.S. IAS IAS IAS

0 0 0.0

(4) Firms switching from U.S. GAAP to IASU.S. U.S. U.S. U.S. IAS IAS IAS IAS

0 0 0.0

(5) Firms reporting under local GAAP during the entire time periodLocal Local Local Local Local Local Local Local

10,756 61,465 88.4

(6) Other (e.g., more than two switches; switches to other foreign GAAP) – – –

Total 12,171 69,528 100.0

The table reports the number of unique firms, and the number and percentage of firm-years across var-ious IAS adoption patterns for our base sample comprising 69,528 firm-year observations from 30 countriesbetween 1990 and 2005 (see table 1). We create a panel of firms’ accounting standards over time based onthe “accounting standards followed” field in Worldscope (field 07536), and then adjusted the coding forcontradictory information from an extensive review of firms’ annual reports. We manually check that theaugmenting process does not introduce artificial IAS switch years. We require a switch year (denoted by IAS

in panel A) for IAS adopting firms to be included in the serious versus label analyses. For firms with two IASswitch years, we utilize the second switch year in our analyses because our manual review found those to bemore reliable. In this case, we set the earlier IAS firm-year observations to missing.

REFERENCES

AMIHUD, Y. “Illiquidity and Stock Returns.” Journal of Financial Markets 5 (2002): 31–56.ARMSTRONG, C.; M. BARTH; A. JAGOLINZER; AND E. RIEDL. “Market Reaction to the Adoption

of IFRS in Europe.” The Accounting Review 85 (2010): 31–61.


BALL, R. “Infrastructure Requirements for an Economically Efficient System of Public Finan-cial Reporting and Disclosure.” Brookings-Wharton Papers on Financial Service (2001): 127–69.

BALL, R. “IFRS: Pros and Cons for Investors.” Accounting and Business Research, InternationalAccounting Policy Forum (2006): 5–27.

BALL, R.; S. P. KOTHARI; AND A. ROBIN. “The Effect of International Institutional Factors ofProperties of Accounting Earnings.” Journal of Accounting and Economics 29 (2000): 1–51.

BALL, R.; A. ROBIN; AND J. WU. “Incentives Versus Standards: Properties of Accounting Incomein Four East Asian Countries.” Journal of Accounting and Economic 36 (2003): 235–70.

BARTH, M. “Global Financial Reporting: Implications for U.S. Academics.” The Accounting Re-view 83 (2008): 1159–79.

BARTH, M.; W. LANDSMAN; AND M. LANG. “International Accounting Standards and AccountingQuality.” Journal of Accounting Research 46 (2008): 467–98.

BARUCH S.; G. A. KAROLYI; AND M. LEMMON. “Multi-Market Trading and Liquidity: Theory andEvidence.” The Journal of Finance 62 (2007): 2169–200.

BOTOSAN, C., AND M. PLUMLEE. “Assessing Alternative Proxies for the Expected Risk Pre-mium.” The Accounting Review 80 (2005): 21–53.

BRADSHAW, M., AND G. MILLER. “Will Harmonizing Accounting Standards Really HarmonizeAccounting? Evidence from Non-U.S. Firms Adopting U.S. GAAP.” Journal of Accounting,Auditing and Finance 23 (2008): 233–64.

BRADSHAW, M.; S. RICHARDSON; AND R. SLOAN. “Do Analysts and Auditors Use Information inAccruals?” Journal of Accounting Research 39 (2001): 45–74.

BURGSTAHLER, D.; L. HAIL; AND C. LEUZ. “The Importance of Reporting Incentives: Earn-ings Management in European Private and Public Firms.” The Accounting Review 81 (2006):983–1017.

BYARD, D.; Y. LI; AND Y. YU. “The Effect of Mandatory IFRS Adoption on Financial Analysts’Information Environment.” Journal of Accounting Research 49 (2011): 69–96.

CAIRNS, D. International Accounting Standards Survey. LLP Professional Publishing, 1999.CAIRNS, D. International Accounting Standards Survey 2000: An Assessment of the Use of IAS in the

Financial Statements of Listed Companies. Henley on Thames, 2001.CAMPBELL, J. “Understanding Risk and Return.” Journal of Political Economy 104 (1996):

298–345.CHORDIA, T.; R. ROLL; AND A. SUBRAHMANYAM. “Co-Movements in Bid-Ask Spreads and Market

Depth.” Financial Analysts Journal 56 (2000): 23–27.CHRISTENSEN, H.; L. HAIL; AND C. LEUZ. “Mandatory IFRS Reporting and Changes in Enforce-

ment.” Working paper, University of Chicago and University of Pennsylvania, 2012.CHRISTENSEN, H.; E. LEE; AND M. WALKER. “Cross-Sectional Variation in the Economic Conse-

quences of International Accounting Harmonization: The Case of Mandatory IFRS Adop-tion in the UK.” International Journal of Accounting 42 (2007): 341–79.

CLAUS, J., AND J. THOMAS. “Equity Premia as Low as Three Percent? Evidence from Analysts’Earnings Forecasts for Domestic and International Stock Markets.” The Journal of Finance 56(2001): 1629–66.

COVRIG, V.; M. DEFOND; AND M. HUNG. “Home Bias, Foreign Mutual Fund Holdings, and theVoluntary Adoption of International Accounting Standards.” Journal of Accounting Research45 (2007): 41–70.

DASKE, H. “Economic Benefits of Adopting IFRS or US-GAAP: Have the Expected Costs of Eq-uity Capital Really Decreased?” Journal of Business Finance and Accounting 33 (2006): 329–73.

DASKE, H., AND G. GEBHARDT. “International Financial Reporting Standards and Experts’ Per-ceptions of Disclosure Quality.” Abacus 42 (2006): 461–98.

DASKE, H.; L. HAIL; C. LEUZ; AND R. VERDI. “Mandatory IFRS Reporting Around the World:Early Evidence on the Economic Consequences.” Journal of Accounting Research 46 (2008):1085–142.

DECHOW, P.; W. GE; AND C. SCHRAND. “Understanding Earnings Quality: A Review of the Prox-ies, their Determinants and their Consequences.” Journal of Accounting and Economics 50(2010): 344–401.


DECHOW, P.; R. SLOAN; AND A. SWEENEY. “Detecting Earnings Management.” The AccountingReview 70 (1995): 193–225.

DING, Y.; O.-K. HOPE; T. JEANJEAN; AND H. STOLOWY. “Differences Between Domestic Account-ing Standards and IAS: Measurement, Determinants and Implications.” Journal of Accountingand Public Policy 26 (2006): 1–38.

DJANKOV, S.; R. LA PORTA; F. LOPEZ-DE-SILANES; AND A. SHLEIFER. “The Law and Economics ofSelf-Dealing.” Journal of Financial Economics 88 (2008): 430–65.

DOIDGE, C.; G. A. KAROLYI; AND R. STULZ. “Why Are Foreign Firms Listed in the U.S. WorthMore?” Journal of Financial Economics 71 (2004): 205–38.

EASTON, P. “PE Ratios, PEG Ratios, and Estimating the Implied Expected Rate of Return onEquity Capital.” The Accounting Review 79 (2004): 79–95.

EASTON, P., AND S. MONAHAN. “An Evaluation of Accounting-Based Measures of Expected Re-turns.” The Accounting Review 80 (2005): 501–38.

ELTON, E. “Expected Return, Realized Return and Asset Pricing Tests.” The Journal of Finance54 (1999): 1199–220.

FOERSTER, S., AND G. A. KAROLYI. “The Effects of Market Segmentation and Investor Recogni-tion on Asset Prices: Evidence from Foreign Stocks Listing in the United States.” The Journalof Finance 54 (1999): 981–1013.

GEBHARDT, W.; C. LEE; AND B. SWAMINATHAN. “Toward an Implied Cost of Capital.” Journal ofAccounting Research 39 (2001): 135–76.

HAIL, L., AND C. LEUZ. “International Differences in the Cost of Equity Capital: Do Legal Insti-tutions and Securities Regulation Matter?” Journal of Accounting Research 44 (2006): 485–531.

HAIL, L., AND C. LEUZ. “Cost of Capital Effects and Changes in Growth Expectations AroundU.S. Cross-Listings.” Journal of Financial Economics 93 (2009): 428–54.

HAIL, L.; C. LEUZ; AND P. WYSOCKI. “Global Accounting Convergence and the Potential Adop-tion of IFRS by the U.S. (Part I): Conceptual Underpinnings and Economic Analysis.” Ac-counting Horizons 24 (2010): 355–94.

HRIBAR, P., AND C. NICHOLS. “The Use of Unsigned Earnings Quality Measures in Tests ofEarnings Management.” Journal of Accounting Research 45 (2007): 1017–53.

HUNG, M. “Accounting Standards and Value Relevance of Financial Statements: An Interna-tional Analysis.” Journal of Accounting and Economics 30 (2001): 401–20.

KARAMANOU, I., AND G. NISHIOTIS. “Disclosure and the Cost of Capital: Evidence from the Mar-ket’s Reaction to Firm Voluntary Adoption of IAS.” Journal of Business Finance and Accounting36 (2009): 793–821.

KOTHARI, S. P. “Capital Markets Research in Accounting.” Journal of Accounting and Economics31 (2001): 105–231.

KOTHARI, S. P.; A. LEONE; AND C. WASLEY. “Performance Matched Discretionary Accrual Mea-sures.” Journal of Accounting and Economics 39 (2005): 163–97.

KYLE, A. “Continuous Auctions and Insider Trading.” Econometrica 53 (1985): 1315–35.LAMBERT, R.; C. LEUZ; AND R. VERRECCHIA. “Accounting Information, Disclosure, and the Cost

of Capital.” Journal of Accounting Research 45 (2007): 385–420.LANDSMAN, W.; E. MAYDEW; AND J. THORNOCK. “The Information Content of Annual Earnings

Announcements and Mandatory Adoption of IFRS.” Journal of Accounting and Economics 53(2012): 34–54.

LANG, M.; K. LINS; AND M. MAFFETT. “Transparency, Liquidity, and Valuation: InternationalEvidence on When Transparency Matters Most.” Journal of Accounting Research 50 (2012):729–74.

LANG, M.; K. LINS; AND D. MILLER. “ADRs, Analysts, and Accuracy: Does Cross Listing in theUnited States Improve a Firm’s Information Environment and Increase Market Value?” Jour-nal of Accounting Research 41 (2003): 317–45.

LANG, M.; K. LINS; AND D. MILLER. “Concentrated Control, Analyst Following, and Valuation:Do Analysts Matter Most When Investors Are Protected Least?” Journal of Accounting Research42 (2004): 589–623.

LANG, M., AND R. LUNDHOLM. “Corporate Disclosure Policy and Analyst Behavior.” The Account-ing Review 71 (1996): 467–92.


LA PORTA, R.; F. LOPEZ-DE-SILANES; A. SHLEIFER; AND R. VISHNY. “Legal Determinants of Exter-nal Finance.” The Journal of Finance 52 (1997): 1131–50.

LESMOND, D.; J. OGDEN; AND C. TRZCINKA. “A New Estimate of Transaction Costs.” Review ofFinancial Studies 12 (1999): 1113–41.

LEUZ, C. “IAS Versus US GAAP: Information Asymmetry-Based Evidence from Germany’s NewMarket.” Journal of Accounting Research 41 (2003): 445–72.

LEUZ, C.; D. NANDA; AND P. WYSOCKI. “Earnings Management and Investor Protection: AnInternational Comparison.” Journal of Financial Economics 69 (2003): 505–27.

LEUZ, C., AND R. VERRECCHIA. “The Economic Consequences of Increased Disclosure.” Journalof Accounting Research 38 (2000): 91–124.

LEUZ, C., AND P. WYSOCKI. “Economic Consequences of Financial Reporting and DisclosureRegulation: A Review and Suggestions for Future Research.” Working paper, University ofChicago, 2008.

OHLSON, J., AND B. JUETTNER-NAUROTH. “Expected EPS and EPS Growth as Determinants ofValue.” Review of Accounting Studies 10 (2005): 349–65.

SLOAN, R. “Do Stock Prices Fully Reflect Information in Accruals and Cash Flows About FutureEarnings?” The Accounting Review 71 (1996): 289–315.

STULZ, R. “Globalization, Corporate Finance and the Cost of Capital.” Journal of Applied Corpo-rate Finance 12 (1999): 8–25.

TAN, H.; S. WANG; AND M. WELKER. “Analyst Following and Forecast Accuracy After MandatedIFRS Adoptions.” Journal of Accounting Research 49 (2011): 1307–57.

VERRECCHIA, R. “Essays on Disclosure.” Journal of Accounting and Economics 32 (2001): 91–180.WATTS, R., AND J. ZIMMERMANN. Positive Accounting Theory. Edgewood Cliffs, NJ: Prentice Hall,

1986.WELKER, M. “Disclosure Policy, Information Asymmetry, and Liquidity in Equity Markets.” Con-

temporary Accounting Research 11 (1995): 801–27.WYSOCKI, P. “Discussion of Ultimate Ownership, Income Management, and Legal and Extra-

Legal Institutions.” Journal of Accounting Research 42 (2004): 463–74.YU, F. “Analyst Coverage and Earnings Management.” Journal of Financial Economics 88 (2008):

245–71.

adopting a label: heterogeneity in the economic consequences around ias/ifrs adoptions

Documents